Home

THE MESONH USER'S GUIDE

Contents

1. HHH HHH COMMENTS SERRE EEE EH EE a E TERETERERERERERERERESERESERESE RETE RETE SERE RETE RETE HERA HARARE HHH E TERETERETERERERERERERERE SERE SE RETE RETE RETE REESE SERERE RE REREREREREAEREAE TETTE TEE ETETETETETETETETEIEREHEHERENERERERERERERERE E HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH CSTORAGE_TYPE MT CMY_NAME COPTE 1 NEWIC 007 CDAD_NAME NIMAX 160 NJMAX 1 NKMAX 44 LCARTESIAN T LTHINSHELL T XBETA 0 XLATO 43 28999999999996 x mesh 2000 y mesh 2000 XZHAT 1 220 49000 XZHAT 2 0 00000 XZHAT 3 220 49000 217 375 gt 218 HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH THEIHE HHH HHH THEIHE HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH HHH THEIHE HHH KTYPE XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZHAT XZH
2. VI CFL compute the CFL number n PURPOSE The purpose of this routine is to compute the CFL number and to write a warning when its value is greater than 1 METHOD 1 The maximum value U of the horizontal wind is computed Then the CFL MW number is computed according to 5 CFL U DeltaT DeltaX M If this number is greater than 1 a warning message is written on the de OUTPUT file together with the instant 1 EXTERNAL 1 Functions MXF MYF computes the averages along the X and Y directions 1 Subroutine FMLOOK to recover the Logical unit associated to a given File 7 4 AN ILLUSTRATIVE EXAMPLE OF A SOURCE COMPILATION 205 1 IMPLICIT ARGUMENTS 11 L2D in module MODD CONF switch for 2D configuration 11 JPHEXT JPVEXT in module MODD PARAMETERS number of external points r1 along the horizontal and the vertical directions 1 REFERENCE J Stein Meteo France Ik 0 DECLARATIONS USE MODI SHUMAN USE MODE TIME USE MODE FM USE MODD CONF USE MODD PARAMETERS IMPLICIT NONE I 0 1 Declarations of dummy arguments REAL DIMENSION INTENT IN PUT PVT 2 components of the horizontal Wind at t REAL DIMENSION INTENT CIN PDXHAT Stretching in x direction REAL DIMENSION INTENT IN PDYHAT Stretching in y direction REAL INTENT IN PTSTEP time step CHA
3. activates the formation of rain by the warm microphysical pro cesses e CPRISTINE ICE Pristine ice crystal type PLAT plates 62 CHAPTER 4 PERFORM A MESONH SIMULATION COLU columns BURO bullet rosettes e LSEDIC Cloud droplets are allowed to sediment when it is TRUE 4 2 10 Namelist NAM TURB CLOUD mixing length for clouds Fortran type default value NMODEL CLOUD integer 999 CTURBLEN_CLOUD 4 characters DELT XCOEF AMPL SAT real 5 XCEL MIN real 0 001E 6 XCELMAX real 0 01E 6 e NMODEL CLOUD model number where the modification of the mixing length in the clouds is computed e CTURBLEN CLOUD type of turbulent mixing length in the clouds IBL89 DELT DEAR see CTURBLEN for meaning e XCOEF AMPL SAT saturation of the amplification coefficient e XCEIL MIN minimum threshold for the instability index in kg kg m s beginning of the amplication e XCELMAX maximum threshold for the instability index in kg kg m s beginning of the saturation of the amplification Diagnostics quantities are written on every synchronuous files mixing length in clear sky mixing length modified amplification coefficient if LTURB_DIAG TRUE in NAM TURBn 4 2 11 Namelist NAM ADV n scalar advection schemes of model n Fortran type default value CUVW ADV SCHEME 6 characters CENATH CMET ADV SCHEME 6 characters FCT2ND CSV ADV SCHEME 6 characters
4. ewer rw sse a 4 3 7 LES averaged fields LLES UPDRAFT TRUE field notation in the dim general comments diac file conditions 7 gt UPFRAC ai _ wpdvaft fraction zt __ from now on computations lt gt UP TH zt made only on the cartesian mask 106 CHAPTER 4 lt a gt Pm lt 1 P SP Sy WA gt um lt Te gt up J X Tr Sup lt Ti gt up lt lt Th up LIUM lt 0 7m lt 6 e 880 gt up 616 gt lt WO gt up UP_WTH 93 gt lt n Sy a BDE 55 gt Em gt UP TVRI lt gt TP TLSV P TVSV lt 0 Sy gt P_WSV aa aiaia P RI P RS JP_TH2 P_THL2 aja FUERC2 fa PERFORM MESONH SIMULATION dim general comments conditions field notation in the diac file r at r r 2 6 pat pat ore zt S Ef zat Pat oom zt om pat oom pat ore pat oom kh EXN EUN t re t mr g h Sy Te ett pect v r zt zt r zt 7 6 r r zt r zt r zt Zt 7 6 EN pat n pat om mmn 0 mtn mtn Tes mn 1 1 1 4 3 LES DIAGNOSTICS 4 3 8 LES averaged fields LLES DOWNDRAFT TRUE field notation in the dim general comments d
5. 2 0 In this case the stratification is taken into account in the Meso NH model in the flottability term The typical length on which this stratification varies is much greater than the domain heigth and the variation can be therefore neglected FALSE The reference anelastic state varies with the altitude e LPERTURB Switch to add a perturbation on the initially horizontally homogeneous fields This perturbation is not balanced 3 perturbation types are implemented in the routine set perturb f90 a spherical perturbation on the dry potential temperature and the moisture fields typical for convection initialization a perturbation on the horizontal components of the wind derived from a streamfunc tion typical for large scale studies This insures the wind against becoming divergent a perturbation on the dry potential temperature field at the first mass level near the ground corresponding to a white noise uniform amplitude in the spectral space This type of perturbation is used in Large Eddy Simulations initialization When TRUE the parameters for the precise definition of the perturbation can be set in the namelist NAM PERT PRE or sometimes must be modified directly in the subroutine set perturb f90 e LFORCING Switch to specify forcing sources When TRUE the precise definition of the forcing is set in the free format part of PRE_IDEA1 nam see 3 3 4 3 2 THE INPUT THE
6. IDEA1 NAM FILE 25 3 2 4 Namelist NAM CONFn configuration variables for modeln Fortran type default value LUSERV logical LUSERC logical LUSERI logical NSV_USER integer see 3 3 2 for more details for these cases e LUSERV Switch to write r vapor mixing ratio in initial file It is reset to TRUE when CIDEAL RSOU or CSTN This has been done in order to avoid to treat the dry case as a particular case but as a moist case with humidity equal to 0 e LUSERC Switch to write re cloud mixing ratio in initial file This case is only allowed when CIDEAL RSOU radiosounding case and KIND PUVTHDMR or KIND ZUVTHLMR e LUSERI Switch to write r ice mixing ratio in initial file This case is only allowed when CIDEAL RSOU radiosounding case and KIND PUVTHDMR e NSV_USER number of scalar variables Note that if NSV_USER is different from 0 the Scalar Variables are initialized to 0 by the program 3 2 5 Namelist NAM DIMn PRE contains dimensions Fortran type default value NIMAX integer 10 NJMAX integer 10 e NIMAX number of mass points in x direction of the initial fileis NIMAX 2JPHEXT JPHEXT corresponds to the number of marginal points in the horizontal directions and is fixed to 1 for the present Meso NH version e NJMAX number of mass points in y direction of the physical domain The total size of the array written in the initial file is NJMA
7. U WSV NSG ADVR neglected advection by resolved flow BU WSV DPGW neglected dyn prod terms terms due to hor 8 gradients BU WSV NSG DPGT other neglected dyn prod terms LE 126 CHAPTER 4 PERFORM A MESONH SIMULATION 4 4 8 Budget of total scalar variance A l terms of the equation of lt y gt lt s gt are computed and stored in the diachronic group BU SV2 comments made for the total equation are valid here ADVM ADVR DPM 0 u lt 82 gt lt gt 5 lt SP lt 81 gt 2 lt 8 gt ps n 2 lt 5 gt lt u s gt s DISS DPR TR ADV DP Sos lt lt gt lt gt lt 4 gt 0 gt 3 Xa 5 OX a PD lt 445 gt 2 lt 8 8 gt TES Ira C TR SBGT field notation in processus dim comments diac file name BU_SV2 SBG DP M dyn prod by mean gradient 2 BU SV2 SBGDPR dyn prod by resolved fluctuations BU_SV2 SBG TR subgrid turbulent transport BU SV2 SBG DISS dissipation 55 2 RESI 44 BUDGET OF RESOLVED SUBGRID TURBULENT QUANTITIES 127 field notation in processus dim comments opposite of tendency of lt gt BU SV2 RES ADV z t n advection Lal E BU SV2 RES FORC z t n advection by large scale
8. 162 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES b two layers are defined with constant stretching in each of these the grid mesh sizes being given near the ground and at top of the model It is possible that the top grid size is never reached if the number of points is not enough for the prescribed stretchings The type of grid YZGRID TYPE is also set to FUNCTN wa the levels are given by the user The type of grid YZGRID_TYPE is set to MANUAL in the namelist and only the number of levels NK MAX is also used in it d only available when the atmospheric input file is MESONH file The levels in the output MESONH file are the same as in the input MESONH file The type of grid YZGRID TYPE is set to SAMEGR for same grid and NKMAX is not specified e only available when the atmospheric input file is a MESONH file The physical levels of the output MESONH file are the same as the lower NKMAX physical levels in the input MESONH file The type of grid YZGRID_TYPE is set to SAMEGR for same grid and NKMAX is specified The variables of this namelist are LTHINSHELL logical FALSE NKMAX integer 60 if HATMFILETYPE GRIBEX integer same as in input file if HATMFILETYPE MESONW YZGRID_TYPE character len 6 FUNCTN if HATMFILETYPE GRIBEX SAMEGR if ZDZGRD real 300 m ZDZTOP real 300 m ZZMAX_STRGRD real Om ZSTRGRD real 0 96 ZSTRTOP real 0
9. 8 S Q 1 field kele dl lr y 2 9 ly ccly 2 9 T ly e ted x y gt er rd e dl x y ly gt dL x y gt Rud ly y gt ete dl z y l gt e wq nat 10 Soe al h a ly ls y gt e tdl lr y gt lz 41 z y 1 gt erg el mop vis z lz y gt la dl x y ly gt gt la qi gt e dl gt a r dl gt y gt c y oe L 2 lz y gt le dl A xr J lt y f x y ly gt e hd N N 48481 AIALA ry Sty Sap Sr 8 3 we je je je SAPNA VIVIVIV SE ho Hi 9 1 EO SISS AJA gu a Sa lulu KS ho Sy EO x ho 1 EO S Qi Io h 1 SSS 1S Sr 8 S SJS gu gu 8 D e Ib d Di S e WN So o Hi ho E x e tke ho 4 E E B equ DID o iky e ho 4 EO AJA D 2 9 1 T ik 2 E EH oN Se J Ree mb D N S E TIETE pe 2 N 454 EB AA e See qe
10. OUTHOST supc to put the output FM files directory name starting at HOME if it begins with home at workdir if it begins with work OUTDIR work INOUTFILES global variables used in prepmodel DEBUGSCRIPT ON OFF ENVIRONMENT SILENTINTERACTIVE INTERACTIVE FILE tosupcrc 45 AN ILLUSTRATIVE EXAMPLE OF A MESONH SIMULATION 131 48 bin sh default input variables for tosupc default values for the NQS parameters N Asencio 13 12 94 specific variables filename which contains the script to be run on remote host JOBFILE 11 0 jobname JOBNAME 12 basename JOBFILE tosupc will not work unless you correctly fill TIME MEM TPN NBP time in seconds for J2 amp J3 jobs TIME memory ex 2000 2Gb 128Mb tori less than 128Gb per node be careful reduce the memory for multi tasks jobs MEM tasks per node only for prep ideal case run or diag number of used CPUs from 1 to 8 The most important for parallel execution TPN 1 mono or multi node on NEC 1 to 4 and on IBM NBP 1 THEHHHHHHHHHEHHHHEHHE tosupc will not work unless you correctly fill TIME MEM TPN NBP Sending a mail abort beg and or at the end job begend beg end no default is abort no is no mail at all MAIL global variables used in tosupc DEBUGSCRIPT OFF SENVIRONMENT SILENTINTERACTIVE SILENTINTERACTIVE BATCH IN
11. ends by 000 51 52 CHAPTER 4 PERFORM A MESONH SIMULATION 4 2 The input EXSEG n nam file We now describe in the following subsection the different NAMELISTs present in a complete EXSEG n nam file Each variable present in a Namelist of the EXSEG n nam file belongs to a declarative module whose name is related to the NAMelist name NAM xxxxx MODD_xxxxx The documentation of the declarative modules MODD_xxxxx is present in the Fortran code and contains a description of each variable present in the Namelist NAM xxxxx Thus we only give the list of the subset of MODD xxxxx present in the Namelist NAM xxxxx with a short description of each parameter In EXSEG n nam file the order of namelists is free not all the free parameters need to be fixed then unset namelists can be ommited For instance if no value is present for the variable CPRESOPT in the NAMELIST NAM DYNn of EXSEG2 nam the index 2 is for model 2 the model will take the value present in the MESONH file used to initialize the model 2 for this segment This information is present in the descriptive part of the MESONH file see Chapter 3 If it is also absent from the MESONH initial file the model will use the default value written in the Fortran code of the model We now list the namelists e Firstly namelists common to all the nested models without suffix n are presented al phabetically from 4 2 1 e Secondly namelists relative to one model with suffix
12. 23 YLBLCH 6 8 067 IF YLINE 4 4 24 YLBLCH 6 8 107 IF YLINE 4 4 25 YLBLCH 6 8 120 ENDIF PRINT TRIM YLINE YLBLCH END IF IF YLINE 1 3 NIL THEN BACKSPACE IUNIT 6 3 VARIABLES AVAILABLE IN THE OUTPUT DIACHRONIC FILE READ IUNIT 24 YLINE 1 4 INIL END IF END DO 99 CLOSE IUNIT 24 FORMAT A4 18 26 FORMAT A6 I8 27 FORMAT AT F4 0 PRINT NBYTE NBYTE IXSIZE IXSIZE IYSIZE IYSIZE PRINT TRIM YMEM YFNAM TRIM YRAC raw OPEN IUNIT FILE YFNAM FORM UNFORMATTED amp access direct RECL NBYTE IOSTAT IO IF 10 0 THEN PRINT PROBLEM WHILE OPENING STOP ELSE PRINT OPEN TRIM YFNAM ENDIF ZVALMIN 999 ZVALMAX O OPEN IUDIR FILE YDIR FORM FORMATTED verbosity level WRITE IUDIR I1 0 format of obs file ll lat lon alt value WRITE IUDIR A2 11 name of the obs file WRITE IUDIR A12 TRIM YOUT name of the obs field WRITE IUDIR A21 YLBLCH unit of the obs field WRITE IUDIR A1 K profile of the obs field WRITE IUDIR A2 2D closing instruction WRITE IUDIR A3 END CLOSE IUDIR OPEN IUOUT FILE YOUT FORM FORMATTED IF NBYTE 1 THEN DO JJ 1 IYSIZE DO JI 1 IXSIZE II II 1 READ IUNIT REC II ZLON ZLONMIN JI ZLONMIN ZLONMAX FLOAT IXSIZE ZLAT ZLATMAX JJ ZLATMAX ZLATMIN FLOAT IYSIZE IVAL ICHAR YTMP1 IF IVAL gt O AN
13. 2D smoothed orography for SLEVE vertical coordinate m Other surface variables are provided through the externalised surface namelists 6 3 1 Some formulae Temperature TEMP The temperature C is computed as 6 1 where is the temperature of the triple point Vapor pressure and Relative humidity The vapor pressure VPRES is computed as P 6 2 1 r Ra R 02 The relative humidity REHU is computed as 2 6 8 0 6 3 When a mixed microphysical scheme is activated during the simulation the saturation vapor pressure is computed over ice at points where temperature is below the triple point 188 CHAPTER 6 COMPUTE DIAGNOSTICS AFTER A MESO NH SIMULATION Refraction coindexes Hill 1980 The refraction coindex COREF is computed as N 77 6 T P 4810 e T 6 e T 6 4 where P and e are in hPa modified refraction coindex MCOREF is computed as M N Z108a 6 5 where Z and a are respectively the altitude and the Earth radius in m Virtual potential temperature THETAV 1 Plo Ra b 6 6 1 7 6 0 where ry is the mixing ratio of total water substance Tw Tod Tec Ted ric Tg d Tg dA TR Equivalent potential temperature THETAE The formulation is taken from Bolton 1980 following its equations 16 21 and 43 3376 0 2 54 ry 1 081r 6 7 Tn 2840 h eeu Intus s t d 0 01
14. NRAD_COLNBR NRAD DIAG NRADTH NRELRC NRELRG NRELRH in namelist NAM BU in namelist NAM BU namelist NAM BU namelist NAM BU namelist NAM BU in namelist NAM BU namelist NAM BU in namelist NAM BU namelist NAM BU in namelist NAM BU namelist NAM BU in namelist NAM BU in namelist NAM BU namelist NAM BU in namelist NAM HU namelist NAM BU in namelist NAM BU in namelist NAM BU namelist NAM BU in namelist NAM BU in namelist NAM BU namelist NAM BU INDEX RRR 88 RRS 90 RRV 86 RSV 91 RTH 85 RU 83 RV 84 RW 84 RRV 86 RTH 85 RU 83 RV 84 RW 84 RR CONF 160 RTH 85 RU 83 RV 84 RW 84 in namelist PARAM RADn 70 in namelist NAM_PARAM_RADn 70 RTH 85 RRC 87 RRG 90 RRH 91 INDEX NRELRI in namelist NRELRR in namelist NRELRS in namelist NRELRV in namelist NRELSV in namelist NRELTH in namelist NRELTKE in namelist NRELU in namelist NRELV in namelist NRELW in namelist NREVARR in namelist NREVARV in namelist NREVATH in namelist NRIMRC in namelist NRIMRG in namelist NRIMRS in namelist NRIMTH in namelist NRIMX in namelist NRIMY in namelist NSEDIRC in namelist NSEDIRG in namelist NSEDIRH in namelist NSEDIRI in namelist NSEDIRR in namelist NAM BU N
15. NVERB 1 GRIDH PRE XDELTAX 5 E2 XDELTAY 5 E2 XHMAX 500 XAX 10 E3 10 NIZS 64 NJZS 2 NEXPX 1 NEXPY 1 LUNITn CINIFILE HYD2D amp NAM_DYNn_PRE CPRESOPT RICHA NITR 4 XRELAX 1 0 48 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE amp NAM_LBCn_PRE CLBCX 1 OPEN CLBCX 2 OPEN CLBCY 1 OPEN CLBCY 2 OPEN amp NAM_VPROF_PRE CTYPELOC IJGRID NILOC 10 NJLOC 2 CFUNU ZZZ CFUNV ZZZ LGEOSBAL FALSE GRn CSURF EXTE amp NAM CH MNHCn PRE LUSECHEM F CSTN 2 285 100000 0 20000 10 10 0 0 40 40 0 01 This file contains the necessary informations necessary to generate the initial conditions for a quasi hydrostatic flow in the weakly non linear regime with a regular vertical grid Now you run the preparative script by prepmodel and you have to answer to the questions asked by the script please enter a name for your simulation directory answer if SIMUL is not initialized Another way to prepare the file outprepideal on local host is to ENVIRONMENT to interactive ENVIRONMENT INTERACTIVE export ENVIRONMENT different parameters set to the values written in the rc files will then appear at the screen and can be modified with your keyboard This configuration is only available on CNRM workstations Then you submit to the remote host for
16. RITOV Radiatif Transfer for Tiros Operational Vertical Sounder RTTOV code version 8 7 PltSatSenBT 2D Brightness temperature K NRTTOVINFO 1 4 nb with Plt Plateforme PltId SatId SenId C Sat Satellite Sen Sensor 1 lt nb lt 10 BTs see section6 3 2 for more details Radiatif transfer code narrow band code Morcrette 1991 SAT_IRBT 2D Brightness temperature in IR channel K CRAD_SAT SAT SAT_WVBT 2D and WV channel K for SAT in METEOSAT GOES E INDSAT with subgrid condensation scheme taken into account default T LRAD_SUBG_COND see section6 3 2 for more details Using of RTTOV is highly recommended to compute brightness temperature 186 CHAPTER 6 COMPUTE DIAGNOSTICS AFTER A MESO NH SIMULATION Variables dim meaning unit MRV MRC MRR 3D Mixing Ratio for Vapor Cloud Rain g kg LVAR_MRW MRI MRS MRG 92 for Ice Snow Graupel Hail g kg R CCCNM CCLOUDM RAINM 3D C2R2 scalar variables g kg 72 MRSVnnn 3D Mixing Ratio for User Scalar Variable n g kg LVAR_MRSV TEMP PRES 3D Temperature C Pressure hPa LTPZH ALT 3D height of model levels geopotentiel in pressure level m REHU VPRES 3D Relative Humidity 96 Vapor Pressure hPa ui COREF MCOREF 3D Refraction coindex modified refraction coindex LCOREF THETAV 3D Virtual potential Temperature K LMOIST V THETAE 3D Equivalent potential Temperature K LMOIST E ABVOR 3D vert
17. 13 init f90 In steps 1 to 6 the user simply takes out from a RCS library the file mod dein f90 In steps 7 to 11 he asks for a second file init f90 which already existed in SIMUL At step 1 mesonh asks for the simulation directory SIMUL this is because the user did not initialize SIMUL before starting his session If SIMUL had been initialized before mesonh would not have asked for it In step 3 prepsource indicates that no name was entered for the user binary library because null character or 0 was entered This can be perfectly correct for example for ad ministrator s tasks or for users who only want to work with the master binary library In our example the user simply hit CR to validate and go on first messages in step 4 shows the user that prepsource has taken the default make file for future compilations because none was existing in the user s SIMUL directory Then prepsource asks for a directory where the source files have to be extracted from Two default values are already set up the second corresponds to the DIRUSER catalog read in prepsourcerc the third to the DIRMASTER catalog which is global to Meso NH The first default will only be set up after the first source file will have been extracted see step 8 The user here enters another directory name in which prepsource will first check for the source file however if it does not find anything it will go on searching in the defaults The file name and ve
18. 6 1 Presentations oe UNS 177 0 2 Execution suom ee 30 UR Se ee i a 177 6 2 1 The prepmodelrc 178 6 2 2 The namelist DIAGI n n 178 6 2 3 Examples of DIAGI nam 180 6 3 Variables available in the output diachronic file 182 6 3 1 Some formulae x ERE CR AE RU avs equ Ee RR A 187 6 3 2 Comparison with satellite 8 190 6 4 References 6 fa she sore Bos XE dE Rp 196 Modify the Fortran sources 197 7 1 Presentation of the different libraries 197 7 2 Meso NH source management 197 7 3 The prepsource procedure 200 7 4 An illustrative example of a source compilation 204 7 5 How to use user s library 211 The MESONH files 213 8 1 The 90 namelists 213 8 2 The Meso NH files 222 26 8d 4 ose o9 4 RUE EGG E RUE Se a 214 8 2 1 The synchronous 214 3 922 The diachronieshle y EE Sd ES 220 8 23 physiographic file 220 8 3 References vec d uy eme Bw Ns ie Ree Eon dE en ew ol 220 6 CONTENTS A il
19. F9801 1 06A12 002 YINIFILE 2 F9802 1 06A12 002 YSUFFIX diag amp NAM_DIAG_ISBAn N2M 2 LSURF_BUDGET T e Namelist file for 6 files using trajectories computation amp NAM_DIAG LVAR_PR T LTOTAL_PR T LTPZH T LVAR_MRSV T LTRAJ T amp NAM_DIAG_FILE YSUFFIX d18 6 YINIFILE 1 NAPE2 1 APEO5 001 YINIFILE 2 NAPE2 1 APE04 001 YINIFILE 3 NAPE2 1 APEO3 001 YINIFILE 4 NAPE2 1 APEO2 001 YINIFILE 5 NAPE2 1 APEO1 001 YINIFILE 6 10 ARP19990919 18 amp NAM STO FILE CFILES 1 2 1 5 001 CFILES 2 NAPE2 1 APE04 001 CFILES 3 NAPE2 1 APEO3 001 CFILES 4 NAPE2 1 APEO2 001 CFILES 5 NAPE2 1 APEO1 001 CFILES 6 APE10_ARP19990919 18 NSTART_SUPP 1 4 NSTART_SUPP 2 2 182 CHAPTER 6 COMPUTE DIAGNOSTICS AFTER A MESO NH SIMULATION 6 3 Variables available in the output diachronic file Atmospheric variables prognostic or diagnostic Pp S Variables dim meaning unit RHODREF 3D Dry density for reference state with orography kg m THVREF 3D Thetav for reference state with orography K Model fields PABSM 3D pression Pa CISO0 PR PREVTK THM 3D potential temperature K CISO TK PRTK UM VM WM 3D wind components m s LVAR RS RVM 3D water vapor mixing ratio kg kg 25 LSUM LSVM LSWM LSTHM LSRVM 3D Large Scale variables LVAR_LS All wind components and wind derivate comp
20. LREFRESH ALL flag to refresh convective columns at every call of the convection scheme LCHTRANS flag to take into account the convective transport for scalar variables can only be used with the options CDCONV KAFR LDEEP switch to use deep convection LSHAL switch to use shallow convection LDOWN switch to use downdrafts in deep convection LSETTADJ switch to allow user defined adjustment time XTADJD user defined deep convective adjustment time if LSETTADJ TRUE X TADJS user defined shallow convective adjustment time if LSETTADJ TRUE LDIAGCONYV switch to store diagnostic variables in module MODD DEEP CONVECTIONn CAPE deep and shallow convective cloud top and base levels up and downdraft mass fluxes NENSM number of additional convective ensemble members for deep convection presently limited to 3 42 THE INPUT EXSEGS N NAM FILE 71 4 2 19 Namelist NAM PARAM RADn options for the radiations of model n Fortran type default value XDTRAD XDTRAD CLONLY CLW CAER CEFRADL CEFRADI COPWLW COPILW COPWSW COPISW real real 4 characters 4 characters 4 characters 4 characters 4 characters 4 characters 4 characters 4 characters XTSTEP XTSTEP RRTM SURF MART SMSH EBCU FOUQ EBCU LCLEAR SKY logical FALSE NRAD COLNBR integer 1000 NRAD DIAG integer 0 XFUDG real 1 It contains the options retained for the radiations scheme used by
21. NCORU 1 NDIFU V 0 NFRCV 0 NCURVV O NCORV 0 NDIFV W 0 NFRCW 0 NCURVW 0 NCORW O NGRAV NADVZTH 1 NFRCTH 0 NPREFTH 1 NDIFTH 1 NCONDTH 1 NHENUTH 0 NDEPSTH 0 NDEF NIMLTTH 0 NBERFITH 0 NCDEPITH 0 0 NADVZTKE 0 NFRCTKE 0 NDIFTKE 0 1 NADVZRV 0 NFRCRV 0 NDIFRV 0 NRELRV 0 NDEPSRV 0 NDEPGRV 0 NCDEPIRV 0 NADVZRC 0 NFRCRC 0 NDIFRC 0 NDCONVRC NDRYGRC 0 NIMLTRC 0 NBERFIRC 0 NCDEF NADVZRR 0 NFRCRR 0 NDIFRR 0 NACCRRR NDRYGRR 0 NGMLTRR 0 NADVZRI 0 NFRCRI 0 NDIFRI O NDCONVRI 220 CHAPTER 8 THE MESONH FILES NHENURI 0 NAGGSRI 0 NAUTSRI 0 NCFRZRI 0 NWETGRI 0 NDRYGRI O NIMLTRI 0 NBERFIRI 0 NCDEPIRI BU RRS LBU_RRS NASSERS 1 NADVXRS O NADVYRS O NADVZRS O NFRCRS O NDIFRS O NSEDIRS O ND NAGGSRS O NAUTSRS O NRIMRS 0 NACCRS 0 NCMELRS 0 NWETGRS 0 NDRYGRS 0 amp NAM BU RRG LBU RRG NASSERG 1 NADVXRG 0 NADVYRG 0 NADVZRG 0 NFRCRG 0 NDIFRG 0 NSEDIRG 0 NI NRIMRG 0 NACCRG 0 NCMELRG 0 NCFRZRG 0 NWETGRG 0 NDRYGRG 0 NGMLTRG O BU LBU_RRH NASSERH 1 NADVXRH 0 NADVYRH 0 NADVZRH 0 NFRCRH NDIFRH BU RSV LBU RSV F NASSESV 1 NADVXSV 0 NADVYSV 0 NADVZSV 0 NFRCSV 0 NDIFSV amp NAM LES NMODNBR LES 1 LTURB LES F NT LES 1000 BLANK XDUMMY1 1
22. OPEN WALL for cyclic open and rigid wall boundary conditions It should be note that CLBCX 1 or CLBCY 1 refers to the lowest index values IJB for X and Y direc tions and CLBCX 2 or CLBCY 2 to the highest index values IIE and IJE Please note that CLBCo 1 CYCL gt 2 These boundaries conditions must correspond to the ones which will be used for the Meso NH run itself e CLBCY array containing 2 elements they represent the type of lateral boundary con dition at the left and right boundaries along y CLBCY 1 and CLBCY 2 respectively They are strings of 4 characters 3 2 12 Namelist LUNIT logical unit names Fortran type default value CINIFILE 28 characters INIFILE e CINIFILE name of the initial FM file which will be produced by PREP IDEAL CASE it will be used as initial file in MESONH numerical simulation 3 2 THE INPUT THE PRE IDEA1 NAM FILE 29 3 2 13 Namelist NAM PERT PRE set analytical perturbations Fortran type default value CPERT KIND characters XAMPLITH real XAMPLIRV real XAMPLIUV real XAMPLIWH real NKWH integer LSET RHU logical XCENTERZ real XRADX real XRADY real XRADZ real e CPERT_KIND Defines the type of the perturbation TH the perturbation is on the thermodynamical fields 0 and r UV the perturbation is on the horizontal wind fields U and V WH the perturbation is a white noise app
23. RES MISC z t ref pressure term radiation microphysics residual of budget of lt 6 gt BU WTHL RESRESI zt 2 82 e lt 0 gt neglected in turb scheme 2 lt w gt c c eee BU WTHL NSG TEND BU WTHL NSG ADVM sink due to subgrid turbulence against 2Ax sponge layer relaxation average from smaller nested models numerical diffusion neglected opposite of tendency of 02 gt neglected advection by mean flow neglected advection by resolved flow BL NN d 118 4 4 4 Budget of total water flux CHAPTER 4 PERFORM MESONH SIMULATION terms of the equation of lt wr gt lt gt are computed and stored in the diachronic group BU WRT comments made for the total Tke equation are valid here ADVM ADVR 2 e E mu m lt gt edo mum USERS lt Uar UWT gt aLa Or Ow lt gt lt uhri gt lt brio gt OXo TP DPR ADV PRES 9 lt wr gt lt oe 2 lt gt lt u lt wr 44 lt gt Xa lt Bb gt dg gt TR T lt p gt Oz DP lt gt lt w gt lt Uw gt lt tar gt OL a La 0 ll i lt 0 gt lt gt Xa SBGT 44 BUDGET OF
24. RESOLVED SUBGRID TURBULENT QUANTITIES 119 field notation in processus dim comments diac file name lt w ce cm BU WRT SBG DPM dyn prod by mean gradient lt 22 527 gt BU WRT SBG DPR dyn prod by resolved fluctuations Ew B BU WRT SBG TR turbulent transport E al correlation term 8 gt thermal production thermal production a BU WRT RES TEND z t opposite of tendency mper HER U WRT RES ADV z t advection BU WRT RES z t advection by large scale BU WRT RES DP zt dyn prod TER BU WRT RES TR transport of resolved E eal BUWRT RES PRES zt presurecomeations BU WRT RESTP zt thermal production BU WRT RES SBGT z t sink due to subgrid turbulence Coriolis toms BUAWRT RES CORT Pat OOS numerical diffusion of lt wr gt RES NUMD z t numerical diffusion against 2A relaxation of lt wry gt E WRT RES RELA sponge layer relaxation 2way nesting of lt wr gt U WEF RES NEST z t average from smaller nested models miscellaneous U WRT RES MISC z t ref pressure term curvature term radiation microphysics residual of budget of lt d gt BU_WRT RES RESI LE ims j lt wr ri gt BU WRT NSG TEND neglected opposite of turb scheme tendency of lt wr gt lt w gt z lt W gt
25. TRUE XRPK 1 polar stereographic projection from south pole 1 gt XRPK gt 0 Lambert projection from south pole XRPK 0 Mercator projection from earth center I XRPK 0 Lambert projection from north pole XRPK 1 polar stereographic projection from north pole XLONORI Longitude in degrees of the origine point not used if LCARTESIAN TRUE This point is the mass point of conformal coordinates x 0 y 0 of the Meso NH grids XLATORI Latitude in degrees of the origine point not used if LCARTESIAN TRUE 3 2 THE INPUT THE PRE IDEA1 NAM FILE 27 3 2 9 Namelist NAM GRIDH PRE horizontal grid definition Fortran type default value XDELTAX real XDELTAY real XHMAX real NEXPX integer NEXPY integer XAX real XAY real NIZS integer NJZS integer e XDELTAX mesh length in meters in x direction on the conformal or cartesian plane It is not used if you read informations in a Meso NH constant file PGD_FILE e XDELTAY mesh length in meters in y direction on the conformal or cartesian plane It is not used if you read informations in a Meso NH constant file PGD_FILE e XHMAX Maximum height in meters hmar for orography case CZS 4 FLAT or ground level for flat orography e NEXPX Exponent exp for orography in case of CZS SINE e NEXPY Exponent exp for orography in case of CZS SINE e XAX Widths in meters a along x for orography in case CZS BELL te
26. Thus the user could have worked interactively before this batch session and then left his source files on SIMUL and run mesonh in batch Some specific remarks concerning prepsource are now made We recall that the input parameters present in prepsourcerc concerning the user s objects BIBUSER etc all are relative to the HOME catalog A 3 Example of the prepmodel script Let us now describe a typical prepmodel session with ENVIRONMENT INTERACTIVE As for prepsource the prepmodel output is preceded by a line number whereas the user s answers are not lod ag FULL SCREEN VISUALISATION OF THE CONTROL PARAMETERS ro the user can change them interactively here again gt 2 gt prepmodel no user library was specified 2 gt Carriage Return alone to continue 2 gt something else and CR to stop gt CR gt user library used 3 gt NMODEL 2 D SUDDEN EEUU an edition of the files get commands is proposed the user may change them 5 gt your job is ready to execute the model s 5 gt execute it alone by tosupc users come fischer mnh master outprepmodel 226 APPENDIX AN ILLUSTRATIVE EXAMPLE OF USE OF PROCEDURES 5 gt or 5 gt add it to users come fischer mnh_master outprepsource 5 gt before executing the global job by tosupc Here step 1 permits the user to change on
27. character 4 LTRANS logical XUTRANS real XVTRANS real LPGROUND FRC logical e LGEOST_UV_FRC switch to use a prescribed geostrophic wind TRUE to integrate a geostrophic wind with a constant Coriolis parameter f 2 x QxSIN XLATO The LCORIO flag of module MODD_CONF must be TRUE FALSE not active e LGEOST_TH_FRC switch to apply a large scale horizontal advection on the potential temperature field The gradients result from the thermal wind balance TRUE to integrate an horizontal advection of 0 FALSE not active e LTEND THRV FRC switch to simulate a large scale 0 and humidity tendency TRUE to integrate a tendency for 0 and ry 58 CHAPTER 4 PERFORM A MESONH SIMULATION FALSE not active e LVERT MOTION FRC switch to simulate large scale vertical transport of all the prognostic fields TRUE to integrate a vertical transport with an upstream scheme FALSE not active e LRELAX_THRV_FRC switch to apply Newtonian relaxation on the potential temper ature and humidity fields TRUE to relax 0 and r towards large scale values FALSE not active e LRELAX_UV_FRC switch to apply a Newtonian relaxation on each horizontal wind component TRUE to relax the horizontal wind towards large scale values FALSE not active XRELAX_TIME_FRC constant damping time for the forced relaxation XRELAX HEIGHT FRC height above which a fo
28. e SSS xn RE amp RB t lt gt esn zt xis esn few x n resin ew SSCS xus eson e _ xo esy zt 3 xw reswa tp xam swen e Precipitation ax INPRRSD zip e SSS Max Precipitation fux MAXINPRRSD zip 7 Evaporation fux EVAPSD p o 3 lt gt RESSvg 3 x RESTHSV feton eass o o eass resavsv o o cans RESUSV atpm SSS x55 atn xus RES WSV 3 ews sws Jat 3 ess swm asp Ss x 4 reswm zb x94 keswa p SSS lt gt r SSS 104 CHAPTER 4 PERFORM A MESONH SIMULATION field notation in dim general comments T 5 4 gt on AJA i i Sus EIER VIVI a a AJA 4 gt mI Sd Se Sf Sf S FY Ly qni So VIV AVA 1 RES WSV2 RES W2SV MENSEM CERERI RE WTLSV if no re replaced by lt 0 08 gt Er amp ViVi viv AIA amp SD VIV amp i i jo a t REWRVSV zipm ma 7 RESTLPZ atp n
29. e s 2 Ib JE Sas SS AJA DIDI Dy Sys KS 2 2 e S l l e ho 4 E en 45 7 Hi y y H lr N gt 2 e 8 e y d Fl y l T o s 2 8 HE S AILAJA A A aJa Ja jsa e je lt e S aja jS e tke IN d D Ibl Sr lt x y Ji lt V y fr lt 0r x y x T le y gt e dl fr lt lx y Tun ar gt e uly dl TL L 2 Ibl 2 2T CHAPTER 4 PERFORM A MESONH SIMULATION notation in the dim general comments ime SLUU dimension 2 5700 h2zt isora SLVV md SLVV hist imaginary SLWW eza parts SJ WW SLUV SUV SLWU SEWV SLTHTH SJ SLTLTL SJ SLWTH SJ WTH SLWTHL SJ WTHL SLRVRV SJ RVRV SLTHRV SJ THRV SI TLRV PSLWRV O Ay 22t o _ SILTHRC SJ_THRC SL TLRO SJ SIL WRC SJ WRC SLRIRI SJ RIRI SL THRI SJ THRI SL TLRI SJ TLRI NN gp pL E e pog pos p EES 071 ec 9 o NNI Ent _ 44 BUDGET OF RESOLVED SUBGRID TURBULENT QUANTITIES 111 field notation in the dim general comments diac file conditions ALT wu Aly SLWRI amp 3zt lt ws y Filey tly
30. gt e vd SJ WRI Ay 2zt r SG AG tly gt dls SISVSV kezin _ lew Angl eT ed SISVSV 1 2 o _ Guy Sr Tey gt dl SCWSV 00 soley ly gt e diy SIWSV _ 4 4 Budget of resolved subgrid turbulent quantities 4 4 1 Budget of total turbulent kinetic energy terms of the equation of Z E e are computed and stored in the diachronic group BU KE Here e and E denote the subgrid and resolved Tke respectively Vr 1 lt gt lt 24 2 2 gt lt E gt lt 247 T 02 gt are main terms of the equations for the horizontal mean of subgrid Tke and resolved Tke in the frame of Boussinesq approximation Note that the computations of the budgets terms are done with the complete equation set and discretization of MESONH The equations here are simplified only for the sake of easier understanding Other terms can arise from the parametrizations of MESONH and will also be taken into account in the budget ADVM ADVR PRES DPM eas ND lt e gt lt ne eee gu ue l Dun o Pug ne lt gt lt gt lt lt u lt uu gt ot s da za lt A 0u lt a gt lt gt lt ule gt lt e gt v Xa se T A DISS DPR TR ADV PRES DP S 1 CN lt ug gt lt gt
31. ice melting Bergeron Findeisen gth cond deposition on ice CCN activation C2R2 or KHKO sedimentation C2R2 or KHKO wet growth of hail Namelist NAM BU RRC budget for cloud water logical integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer PERFORM MESONH SIMULATION Fortran type default value Hy gt b U Or O OO c Oc Fortran type default vale LBU_RRR NASSERR NNESTRR NADVRR NADVXRR NADVYRR NADVZRR NFRCRR NDIFRR NRELRR NACCRRR NAUTORR NREVARR NSEDIRR NSFRRR NACCRR NCFRZRR NWETGRR NDRYGRR NGMLTRR NWETHRR NHMLTRR 42 THE INPUT EXSEGS N NAM FILE budget switch time filter Asselin nesting total advection advection along x advection along y advection along z forcing numerical diffusion relaxation accretion of rain droplets autoconversion into rain droplets rain evaporation sedimentation of rain droplets spontaneous freezing ICE3 or ICEA accretion of rain water or ICE4 conversion freezing ICE3 or ICEA wet growth of graupel ICE3 or ICEA dry growth of graupel ICE3 or ICE4 graupel melting ICE3 or ICEA wet growth of hail ICEA melting of hail ICEA 4 2 37 Namelist NAM BU RRR budget f
32. mem upward mass fux 4 3 4 LES pdf LLES_PDF TRUE field notation in the dim general comments EM CEA URN PDFS PDE TH zipa amon p sequiltothe PDFJV PDEW ztpa mmbeofmaks PDFJ PDE THV apa dimension a is equal to the PDF PDERV ztba mimber of pdf intervals NPDF PDF F PDERC 7 PDF R PDERT i E PDFa PDETHL y E PDFE PDERR y PDFE PDERI fapa o E PDFR PDERS fapa PDFE PDERG 4 3 5 LES averaged fields LLES_ RESOLVED TRUE 102 CHAPTER 4 PERFORM A MESONH SIMULATION field notation in the dim general comments diac file conditions warning contains both turbulent and gravity wave fields mure r lt gt Mesme lt gt _____ 2 r jJ 4 3 LES DIAGNOSTICS 103 field notation in the dim general comments diac file conditions RES_UTH warning contains both turbulent and gravity wave fields RES WTH ae SSS SES ee lt gt zip SSCS lt gt RE amp VRV zip 71 x4 RESWRy xm Resko ip re SSS lt gt RESTARG e SSS s resme tp re tp r SSCS lt gt RESURO tp re SSS xs RESVRO i 1 xw RESWRO tb
33. mw BURE SBG TP zt thermal production BUKE st disipo numerical diffusion of lt e gt BUKE SBG NUMD z t numerical diffusion MEI c relaxation of Ges zt sponge layer relaxation Fmiscellaneous Buke seeme a SY residual of budget of lt e gt BUKE SBGRESI zt must bezero g ES opposite of tendency ofc E gt eee ee by mean flow lt E gt UKE RES FORC 7 6 advection by large scale ww BU KE RES DP dyn prod indi Hi p 2 Foa by itself lt gt PRES presure corcais BUKE RES TP par neri poder lt Bag ug gt UKE RES SBGT Z T sink due to subgrid turbulence Coriolis terms 80 XE RESCORI should be zero for E numerical diffusion of lt E gt BULK RES NUMD Z T numerical diffusion against 2Ax relaxation of lt E gt BU RES RELA sponge layer relaxation E nesting of E gt RES NEST 2 1 average from smaller nested models BUKE KESMISC zi curvature tems residual of budget of lt E BUKE RESRESI zt must bezero SI 114 CHAPTER 4 PERFORM A MESONH SIMULATION Note that if a term is zero because the process accounting for it is not activated in the model the term is not listed in the diachronic file So in order to know which terms have been computed and stored use the command print BU_KE proc in diaprog 4 4 2 Budge
34. user in namelists from the externalized surface facility PGD namelists NAM COVER and NAMLISBA else the PREP_PGD fields are taken into account namelists are then list alphabetically 3 2 1 Namelist NAM BLANK available variables see Perform a MESONH simulation for description 3 2 2 Namelist NAM CH MNHOCn init chemistry scalar variables If you initialize MNH C using PREP IDEAL CASE use the following namelist variables Fortran type default value LCH_INIT_FIELD logical FALSE LORILAM logical FALSE CCHEM_INPUT_FILE 80 characters MNHC input e LCH_INIT_FIELD switch to activate initialization subroutine CH_INIT_FIELD e CCHEM_INPUT_FILE name of the general purpose input file for initialization e LORILAM switch to activate initialization chemical aerosol only if LCH_INIT_FIELD T 3 2 THE INPUT THE IDEA1 NAM FILE 23 3 2 3 Namelist NAM CONF PRE configuration variables Fortran type default value LCARTESIAN logical TRUE LPACK logical TRUE CEQNSYS characters DUR NVERB integer 5 CIDEAL 4 characters CSTN CZS 4 characters FLAT LBOUSS logical FALSE LPERTURB logical FALSE LFORCING logical FALSE e LCARTESIAN Switch for cartesian geometry TRUE for cartesian geometry FALSE for conformal projection e LPACK Switch to compress FM file for 1D or 2D version e CEQNSYS Equation system resolved by the MESONH model LHE Lipps and HEmle
35. BU RRI 89 in namelist NAM BU RRS 90 namelist description 60 namelist description 161 namelist description 51 namelist description 22 namelist description 87 namelist description 90 namelist description 76 NAM CH SOLVERn NAM CONF NAM CONF PRE NAM CONFn NAM_DIMn_PRE NAM_DUST NAM_DYN NAM_DYNn NAM DYNn PRE NAM FILE NAMES NAM FMOUT namelist description 78 namelist description 51 namelist description 23 namelist description 25 61 namelist description 25 namelist description 53 namelist description 54 namelist description 62 namelist description 26 namelist description 156 169 namelist description 91 NAM BU RRI namelist description 89 namelist description 55 NAM FRC namelist description 55 NAM BU RRR NAM BU RRS NAM BU RRV NAM BU RSV NAM BU RTH NAM BU RTKE NAM BU RU NAM BU RV NAM BU RW NAM BUDGET NAM_CH_MNHCn NAM CH ORILAM namelist description 88 namelist description 90 namelist description 86 namelist description 91 namelist description 85 namelist description 86 namelist description 83 namelist description 84 namelist description 84 namelist description 81 namelist description 74 NAM GRID PRE NAM GRIDH PRE NAM GRn PRE NAM HURR CONF NAM LBCn NAM LBCn PRE NAM LES NAM LUNITn NAM MESONH DOM NAM NESTING NAM NUDGINGn NAM PARAM C2R2 namelist description 26 nam
36. BU RTH 91 NAM BU RTH 89 NAM BU RRR 88 NAM BU RTH 90 NAM BU RTH 86 NAM BU RSV 91 NAM BU RTH 85 NAM BU RTKE 86 NAM BU RU 83 NAM BU RV 84 in namelist NADVYRC in namelist NADVYRG in namelist NADVYRH in namelist NADVYRI in namelist NADVYRR in namelist NADVYRS in namelist NADVYRV in namelist NADVYSV in namelist NADVYTH in namelist NADVYTKE in namelist NADVYU in namelist NADVYV in namelist NADVYW in namelist NADVZRC in namelist NADVZRG in namelist NADVZRH in namelist NADVZRI in namelist NADVZRR in namelist NADVZRS in namelist NADVZRV in namelist NADVZSV in namelist NADVZTH in namelist NADVZTKE in namelist NADVZU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU RW 84 RRC 87 RRG 90 RRH 91 RRI 89 RRR 88 RRS 90 RRV 86 RSV 91 RTH 85 RTKE 86 RU 83 RV 84 RW 84 RRC 87 RRG 90 RRH 91 RRI 89 RRR 88 RRS 90 RRV 86 RSV 91 RTH 85 RTKE 86 233 234 in namelist NAM BU RU 83 NADVZV NADVZW NAGGSRI NAGGSRS NAM ADVn NAM AERO CONF NAM BLANK NAM BLANK NAM BU RRC NAM BU RRG NAM BU RRH in namelist NAM BU RV 84 in namelist NAM BU RW 84 in namelist NAM
37. BU_WRT ADVM z t neglected advection by mean flow lt wri gt BU WRT NSG ADVR z t neglected advection by resolved flow terms due to 70 gradients BU WRT DPGW 2 6 neglected dyn prod terms terms due to hor 7T gradients BULWRT NSG DPGT z t other neglected dyn prod terms 120 CHAPTER 4 PERFORM A MESONH SIMULATION 4 4 5 Budget of liquid temperature total water covariance terms of the equation of lt gt lt Ort gt are computed and stored in the diachronic group BU_THLR All comments made for the total Tke equation are valid here ADVM ADVR DPM 7 7 Tul 7 9l lt gt 7 lt 0 gt lt Ori gt lt Ua gt lt Ori gt lt uabr gt lt 0 gt lt gt ot Ut Ut Ika it Oa a t Ox Or 06 lt u d gt lt ri gt lt gt LET TN Des Clt tu Ur DISS DPR TR ADV DP PEN il ae nS TIS _ lt gt i lt Ort gt lt gt Uf lt us 0 gt lt dft 5 Ley 3 5 lt gt FK Oiz 64 fez o0 gt etes Q Q TR SBGT field notation in processus dim comments diac file name lt wh gt n BU THLR SBG DP M z t dyn prod by mean gradient lt wr gt lt 0 gt lt ual gez t gt BU THLR SB
38. CCHEM_INPUT_FILE CCH_TDISCRETIZATION NCH_SUBSTEPS LCH_TUV_ONLINE CCH_TUV_LOOKUP CCH_TUV_CLOUDS XCH_TUV_ALBNEW XCH TUV DOBNEW XCH_TUV_TUPDATE CCH_VEC_METHOD NCH_VEC_LENGTH XCH_TS1D_TSTEP CCH_TS1D_COMMENT CCH_TS1D_FILENAME logical logical logical logical logical logical 80 characters 10 characters integer logical 80 characters 4 characters real real real 3 characters integer real 80 characters 80 characters e LUSECHEM switch to activate chemistry FALSE FALSE FALSE FALSE FALSE FALSE EXSEGI nam SPLIT 1 TRUE PHOTO TUV39 INONE 1 1 600 1000 600 comment TOLD e LCH_INIT_FIELD switch to activate initialization subroutine CH_INIT_FIELD 4 2 THE INPUT EXSEG N NAM FILE 77 LCH SURFACE FLUX switch to activate chemical surface fluxes subroutine CH_SURFACE_FLUX LCH CONV SCAV switch to activate scavenging of chemical species gazeous or aerosol and dusts by convective precipitations LCH EXPLICIT SCAV not yet implemented LCH_CONV_LINOX switch to activate the production of NOx by LIghtning flashes inside deep convective clouds and its transport LCHTRANS must be set to TRUE If LUSECHEM F scalar variable named LINOX are written in the LFI file LUSECHEM T the convective source is added to the NO chemical variable CCHEM INPUT FILE name of the general purpose input file CCH TDISCRETIZATION temporal discretizatio
39. E 2 XDUMMY2 O E 0 XDUMMY3 4000 XDUMMY4 11 XDUMMY5 0 3000000000000007 XDUMMY6 XDUMMY7 0 0 XDUMMY8 0 0 NDUMMY1 0 NDUMMY2 0 NDUMMY3 0 NDUMMY4 0 NDUMMY5 0 NDUMMY6 NDUM NDUMMY8 0 LDUMMY1 LDUMMY2 T LDUMMY3 T LDUMMY4 T LDUMMY5 T LDUMMY6 T LDUMMY7 T LDUMMY8 T CDUMMY2 0 0 NVTURBSV 0 N CDUMMY3 CDUMMYS CDUMMY6 CDUMMY7 CDUMMY 5 amp NAM FRC LGEOST UV FRC LGEOST TH FRC LADVEC RV FRC LVERT MOTION FRC LRELAX THRV FRC F LRELAX XRELAX TIME FRC 10800 XRELAX HEIGHT FRC 0 CRELAX HEIGHT TYPE FIXE LTRANS T XUTRANS 13 XVTRAN amp NAM CH SOLVER CSOLVER SIS NSSA 0 NSSAINDEX 1000 0 XRTOL 1 E 3 XATOL 0 1000000000000001 NRELAB 2 NPED 1 NMAXORD 5 LPETZLD T CMETHOD N CNORM A NTRACE 0 XALPHA 100 XFAST 0 1000000000000001 NQSSAITER i XDTMIN 0 1000000000000001 XDTMAX 600 XDTFIRST 10 amp NAM PARAM ICE LWARM T CPRISTINE ICE PLAT CHEVRIMED ICE GRAU CDUMMY4 8 2 2 The diachronic file It is a file obtained during a segment of simulation or resulting of the conversion of a synchronous file with conv2dia for graphical purposes The file directly obtained during the simulation has a name ended by 000 and contains records such as averaged variables tendancies fluxes stored at different times of the simulation on the whole or so
40. ECMWF T5 climatology NONE no aerosol e CEFRADL liquid effective radius calculation MART based on Martin et al 1994 JAS C2R2 based on the prediction of the number concentrations Recommended for the 2 moment microphysical schemes PRES very old parametrization as f pressure simple distinction between land 10 and ocean 13 e CEFRADI ice water effective radius calculation ice particle effective radius from Liou and Ou 1994 SURT ice particle effective radius f T IWC from Sun and Rikus 1999 C3R5 based on the prediction of the number concentrations Recommended for the 2 moment microphysical schemes not yet available for mixed clouds 40 fixed 40 micron effective radius COPWLW cloud water LW optical properties SMSH Smith Shi formulation SAVT Savijarvi formulation only recommended with 2 moment microphysical schemes with small precipitation COPILW ice water LW optical properties EBCU Ebert Curry formulation Smith Shi formulation only with CLW RRTM FULD Fu Liou formulation only with CLW MORC COPWSW cloud water SW optical properties FOUQ Fouquart 1991 formulation SLIN Slingo 1989 formulation COPISW ice water SW optical properties 42 THE INPUT EXSEGS N NAM FILE 73 EBCU Ebert Curry formulation FULIP Fu Liou formulation e LCLEAR SKY When this logi
41. Execution DIAG is run with the Meso NH procedure prepmodel used with the namelist DIAG1 nam example of the namelist file is given in section 6 2 2 Up to 24 FM files can be treated identically in a single job Two main tasks are performed by the procedure prepmodel 1 After the get of the input FM file diagnostics are computed by the routine write_lfifm1_for_diag Diagnostic variables as well as some prognostic variables are written in a temporary FM file by call to the physic parametrisations monitor phys param1 by the routine write lfifm1 for diag supp Diagnostic variables related to convection radiation and surface computed in the physic parametrisations monitor are written here see the graphic documentation by J Duron on the Meso NH web site 177 178 CHAPTER 6 COMPUTE DIAGNOSTICS AFTER A MESO NH SIMULATION 2 Each temporary FM file can be converted into a diachronic file by the program conv2dia see TRAITEMENT GRAPHIQUE DES FICHIERS SYNCHRONES produits PAR LE MODELE MESONH J Duron called by the script transfer x from prepmodel or kept in FM format and renamed These output files are then stored in a workspace in the super computer or in the archive machine or transferred to your workstation 6 2 1 The prepmodelrc file In the file prepmodelrc the input and output host directories and login control variables refer to the input and output files The other control variables to initialize sp
42. HPGDFILE PGDFILE 10km CINIFILE CPL_example2 amp NAM REAL CONF NVERB 5 amp NAM VER GRID NKMAX 10 YZGRID TYPE MANUAL ZHAT 0 1050 2100 3250 4300 5200 6100 7000 8000 9000 10000 5 8 HORIZONTAL INTERPOLATION FROM A MESO NH FILE SPAWNING 169 5 8 Horizontal interpolation from a MESO NH file SPAWN ING This program performs the horizontal interpolation from one MESO NH file into another re spectively file 1 and file 2 The grid of the file 2 must be exactly included in the grid of file 1 The file 2 can be directly used for a model run but it contains smooth surface fields expecially the orography It is possible to run the model with the two files with gridnesting interaction since a iterative procedure insures the gridnesting condition on the orographies domain of the file 2 can be defined either 1 by namelist NAM GRID2 SPA specification 2 with the domain of another FM file which grid is coherent with the input file For example this file can be a file created by PREP PGD with a domain defined from the domain of file 1 and the same type of specifications as those in NAM_GRID2_SPA see above 5 8 1 The prepmodelrc file SPAWNING is run with the procedure prepmodel The input and output host directory and login control variables refer to all the input and output files other control variables to initialize specifically are e MAINPROG SPAWNIN
43. LWF DOWN CS idem for Clear Sky results LWF UP CS LWF NET CS SWF NET CS 2 DTRAD SW CS DTRAD LW CS di RADSWD VIS CS RADSWD NIR CS RADLWD CS 69 PLAN ALB VIS PLAN ALB NIR 2D planetary albedo NRAD 3D 3 4 5 PLAN TRA VIS PLAN_TRA_NIR 2D planetary transmission 59 PLAN ABS VIS PLAN ABS 2D planetary absorption e EFNEB DOWN EFNEB_UP 3D equivalent emissivity Morcrette scheme NRAD_3D 4 5 FLWP FIWP 3D liquid ice water path g m 5 EFRADL EFRADI 3D cloud liquid water and ice effective radius um SW_NEB RRTM LW NEB 3D _ effective cloud fraction OTH VIS 1 OTH NI2 OTH NI3 3D optical thickness SSA VIS 55 1 SSA_NI2 SSA_NI3 3D _ single scattering albedo ASF VIS ASF_NIR1 ASF_NIR2 ASF_NIR3 3D _ asymetry factor O3CLIM 3D Pa Pa climatological ozone content CUM_AER_LAND CUM_AER_SEA 3D _ cumulated optical thickness CUM_AER_DES CUM_AER_URB gt ofthe different aerosols CUM_AER_VOL CUM_AER_STRB from the top of the domain Dust variables LDUST in YINIFILE des DSTMOnM 3D Dust 0 order moment of the lognormal mode n ppbv DSTM3nM 3D Dust 3 order moment of the lognormal mode n ppbv DSTM6 nM 3D Dust 6 order moment of mode n if LVARSIG ppbv DSTRGAn 3D Dust number mean Radius of the lognormal mode n um DSTRGAMn 3D Dust Mass mean Radius of the lognormal mode n um DSTNOAn 3D Dust Numbe
44. MESONH user can get part of them directly on its screen He can get brief informations on the script procedure by making script He will then get a list of information purpose synopsis list of input parameters and input files It also contains some words on the calling scripts and the called scripts He can also access a man type information by man script It contains an extended information about the procedure including especially a more or less detailed description of the method used in the script Some informations are similar to those present in the on line help As previously quoted each script except tosupc creates a file called out script This file contains set of executable UNIX commands that can be executed on the remote host Fujitsu NEC or UNIX workstation and every elementary step adds a new part to this file In this picture and later on we will refer to local host as the machine on which the preparative script is run The output file i e out script is executed on the remote host Nevertheless the remote host may be identical to the local host if desired In the first chapters we will use the model as it stands The compilation of the MESONH sources has been previously done and the binaries objects generated during this compilation step have been collected in binary libraries localized on the remote host different bi naries are loaded by the procedure outprep in order to obtain th
45. MESONH model or of a nested simulation interpolation from the Larger Scale MESONH model to a new finer mesh MESONH model The third part is the temporal integration of the model starting from the initial values read in the initial MESONH files During the integration the model outputs are stored in MESONH files identical to the initial file In the fourth part diagnostic variables can be computed after the simulation To perform these different parts some Unix procedures have been developped and are described in this book A lot of parameters are to be given e to the procedure itself as for instance in which directory the input MESONH files are to be searched e to the fortran source to correctly configurate the model as for instance the model time step The presentation of these procedures is realized in Chapter 2 because they are fondamental tools to correctly manipulate the Meso NH atmospheric simulation system We will delay the presentation of how to modify the fortran source to the Chapter 7 where the source manager is presented in detail We will present in Chapter 3 how to generate initial conditions for a Meso NH simulation fields are stored in a file whose structure is described in the first Annexe This file contains the values of all prognostic and some diagnostic fields in binary form This file type will be called a FM file in the following We also describe the file which contain the free parameters of
46. P r ande T 062 r Vorticity quantities The relative vorticity UM1 VM1 WM1 is computed as 2 oi w v gt VAU G oj 72 pes a 02 On v gt 6 8 rE 6 8 The absolute vorticity ABVOR takes into account the rotation of the earth 2 6 9 6 3 VARIABLES AVAILABLE IN THE OUTPUT DIACHRONIC FILE 189 Potential Vorticity POVOM The Ertel potential vorticity is computed as 27 0 550 6 10 Pdref The unit is the Potentiel Vorticity Unit 1PVU 10 K m kg7 s7 Moist Potential Vorticities The virtual potential vorticity POVOV is Bos 6 11 Pdref and the equivalent virtual potential vorticity POVOE eV 0 Ps eV Ge 6 12 Pdref Geostrophic and ageostrophic winds UM88 VM88 UM89 VM89 With the LHE system see book1 chapter 2 the geostrophic wind is computed as Ug _ 1 ACpaP s l 1 ACpdF ores H 6 13 f 8 GE With the MAE and DUR systems see book1 chapter 2 the geostrophic wind is computed as 2 1 1 ug Ug F Colores 6 14 where A pos 2 8 Poo f The ageostrophic wind is computed as Uag U Ug Vag V Vg Mean sea level pressure MSLP The surface pressure is first computed as the mean between the pressure at the first mass level and at the level below Then it is reduced to the mean sea level where the height is zero following the Laplace law Z P
47. PDF MIN Minimum of rain mixing ratio pdf e XRR_PDF_MAX Maximum of rain mixing ratio pdf e XRI PDF MIN Minimum of ice mixing ratio pdf e XRLPDF MAX Maximum of ice mixing ratio pdf e XRS PDF MIN Minimum of snow mixing ratio pdf e XRS PDF MAX Maximum of snow mixing ratio pdf e XRG_PDF_MIN Minimum of graupel mixing ratio pdf e XRG PDF MAX Maximum of graupel mixing ratio pdf e _ MIN Minimum of total mixing ratio pdf e XRT_PDF_MAX Maximum of total mixing ratio pdf e XTHL_PDF_MIN Minimum of 6 pdf e XTHL PDF MAX Maximum of 6 pdf 4 3 LES diagnostics 4 3 1 notations T a mean valne ofa in the grid resolved quantity 3D lt Q gt up horizontal mean value of in updrafts 1D Ce on eii rats than gt ac considered lt gt down horizontal mean value of a in downdrafts 1D tm y pint ith ml Examples 100 CHAPTER 4 PERFORM A MESONH SIMULATION Ssubgrid flux or co variance a m mean value ofa in each grid mesh 3D 0 value of the turbulent fluctuation in cach gri mes 321 mean value in each grid mesh of subgrid flux or co variance resolved flux or co variance in each grid mesh 3D zas horizontal mean value ofa SID 2a gt O horizontal mean value of a sabgiid Hucua on 1 ID x 0 horizontal mean value of a resolved fluctuation 1D Ea horizontal mean value of subgr
48. SIMUL directory If MAINPROG is set to PREP IDEAL CASE then a namelist called EXPRE1 nam is read Again make sure in this case that this file exists and is up to date on SIMUL One of the main jobs in prepmodel is to get and save the Meso NH input and output files on the remote machine on which the job will be executed Thus there is some moving around of information and you have to be aware of it Furthermore for some specific use like debugging you need to know where you can find all you need to run totalview Figure A 2 gives an overview of what kind of files are copied and where Mainly it should be recalled that the code is compiled on TEMP that it is run on TEMP execdir and that the source as well as the executable code is saved on workdir if DEBUG cdbx We stress that the Meso NH output data files are stored via the script transfer x which is called directly by the fortran programs However you need to define carefully the parameters OUTHOST and OUTDIR to make sure that transfer x is saving the files where you expect it We recall also that depending on the dummy argument in FMOPEN the descriptive and part of Meso NH file will be concatenated via cpio or not If file transfers have failed prepmodel will try to save the local data files either on SuperCalc if enough space or on Archive A 4 Example of the tosupc script Let us now look at a typical INTERACTIVE tosupc session conventions are the sam
49. XLES TEMP MEAN START real none XLES_TEMP_MEAN_END real none XLES TEMP MEAN STEP real 3600 s LLES CART MASK logical FALSE NLES IINF integer physical domain boundary JPHEXT 1 NLES ISUP integer physical domain boundary NIMAX JPHEXT NLES_JINF integer physical domain boundary JPHEXT 1 NLES_JSUP integer physical domain boundary NJMAX JPHEXT LLES_NEB_MASK logical FALSE LLES_CORE_MASK logical FALSE LLES_MY_MASK logical FALSE e LLES_MEAN logical switch for computation of the mean vertical profiles of the model variables e LLES RESOLVED logical switch for computation of the mean vertical profiles of the resolved fluxes variances and covariances e LLES SUBGRID logical switch for computation of the mean vertical profiles of the subgrid fluxes variances and covariances e LLES_UPDRAFT logical switch for computation of the updraft vertical profiles of some resolved and subgrid fluxes variances and covariances e LLES DOWNDRAFT logical switch for computation of the downdraft vertical profiles of some resolved and subgrid fluxes variances and covariances 96 CHAPTER 4 PERFORM A MESONH SIMULATION LLES SPECTRA logical switch for computation of the non local diagnostics 2 points correlations and spectra NLES LEVELS list of model levels where the local quantities are computed Default is all model levels per default the vertical profiles are computed on the MESO NH grid XLES HEIGHTS list
50. additional file 300 Mo which contains coefficients and the binary library of the transfer code itself so in the DIAG1 nam file add the following lines DIAG NRTTOVinfo 1 3 6 200 FILESGET_LIST rttov87_rtcoef tar in the prepmodelrc file add the following line LOAD_OPT MESONH binaries librttov8 7 a Here the first number 3 corresponds to the Meteosat platform the second 6 to the satellite Meteosat 6 and the third 20 to the Seviri sensor the fourth 0 is not used To simulate other instruments use the code given in Tables 6 1 and 6 2 reproduced below from the RT TOV users guide see http www metoffice com research interproj nwpsaf rtm rtm rttov8 html RTTOVid NOAA DMSP Meteosat GOES GMS FY 2 ERS EOS METOP ENVISAT MSG FY 1 ADEOS MTSAT CORIOLIS 6 1 Platforms supported by RT TOV 8 7 as at 17 Nov 2005 in normal text Platforms in italics are not yet supported by RTTOV 8 7 but soon will be taken from RTTOV 8 7 Users guide Table 2 page 4 192 CHAPTER 6 COMPUTE DIAGNOSTICS AFTER A MESO NH SIMULATION RTTOVid Sensor Channel RTTOV 8 Channel HIRS 1 to 19 1 to 19 MSU 1 to 4 1 to 4 SSU 1 to 1 to AMSU A 1 to 15 1 to 15 AMSU B 1 05 1 05 AVHRR 3b to 5 1to3 SSMI 1to7 1to4 VTPRI 1 to 8 1 to 8 VTPR2 1 to 8 1 to 8 1 to 9 1 to 9 SSMIS 1 to 24 1 to 21 AIRS 1 to 2378 1 to 2378 HSB 1 to 4 1 to 4 MODIS 1 to 17 1 to 1
51. by mean gradients SV2 RES TR z t n resolved transport of peu em pen to subgrid turbulence numerical diffusion of lt 5 gt BU SV2 RES NUMD numerical diffusion against 2Ax SU BU SV2 RES RELA sponge layer relaxation 2way nesting of lt Sy ETIN RES NEST average from smaller nested models BUSVZ RESMISC ain chemisny E of budget of 5 BUSV2 ztn mutbezro BU SV2 NSG TEND z t n neglected of neglected in turb scheme tendency of lt 82 gt lt w gt lt s gt BU SV2 NSG ADVM z t n neglected advection d ES 5 n CHAPTER 4 PERFORM A MESONH SIMULATION 4 5 An illustrative example of MESONH simulation We still refer to the example for PREP IDEAL CASE for the control of the environment vari ables We will still use the rc files to enter the input parameters for the different procedures prepmodelrc tosupcrc and the file including the F90 namelist is called EXSEG n nam n 1 in this example e The final rc file are given below FILE prepmodelrc 4 bin sh default input variables for prepmodel comfort variables OUTSCRIPT instead of outprepmodel outprepideal outprep_pgd outprepreal outprepspawn outprepdiag SLEEPSECONDS EDITOR vi instead of 10 specific variables Control tools commands in prepmodel global var yes or no default variable TOOLSCONTROL no tori hpce tora fuji u
52. conditions lt tc e bw gt CLTHTH lt a Wry l gt AG 8er CTL baat AG fari re coen b gt CLWTH coen SG CLWTH 224 lt fo l gt CLWTHL 224 wy Bart CLWTHL 1224 erenneren CLRVAV lt eeu re lt A ew role ow gt CLTARV heat lew gt CLTHRV r _ CLTLAV at re 221 221 AJA A DoD l 3 a lt AAA o a Tuy y xy y T lr Y T Y ly Y ly gt a dad il y ly gt gt 124 VIVIVIVIV ALAA DIDI Dy S54 4 a 9 e je Re few J SS em E NI 8 a J eS me S Jz x lz y a BUR S 3 e que ji AIA E e ye ye ye 4 2 1 9 gt gt gt gt gt gt gt Pd gt gt i gt TITIS IS S w ce oz WE S J a gt gt ly 7 m y F ly gt 12 7 Sulz Y F ly e ae 4 z 2 Ty Ty Tc Tc Tc Tc Tc Tc Tc Tc Tij Tij Tij Tij Tij Tij Tij Tij 5 5 5 5 Sy Sr e 110 4 3 11 LES spectra PR EE ALAA
53. data file for the model 3 same projection definition of reso lution domain PREP NEST PGD this program checks all the three PGD files at the same time and imposes the conformity between them extractecmwf or extractarpege it extracts the surface and altitude fields for one date for model 1 The extraction must be done separately for each date and time for the initial file and each of the coupling file of model 1 PREP REAL CASE this program is running several times for the initial file and the coupling files of model 1 MESONH this step is optional If you do not wish to start all the models at the same time you can decide to run the model 1 before the model 2 starts ZOOM PGD this step is optional If you want to start the model 2 on a smaller domain than the one of the PGD file defined at steps 1 and 2 for the model 2 you must use this program SPAWNING when you want to start the model 2 you must use this program to compute the horizontal interpolations from the model 1 to the model 2 It is used only once for the initialisation of model 2 PREP REAL CASE It is used only once to compute the initial file for the model 2 Do not change the vertical grid MESONH again this step is optional If you do not wish to start model 3 at the same time as model 2 you can decide to run the models 1 and 2 alone before ZOOM PGD again this step is optional If you want to start
54. finally decides not to perform the new release he intended to do when making nh extract he can undo this extraction nh unextract source f90 MYOWNL In this case no advance is performed in the release numbers The Figure 1 presents picture illustrating all the actions that have been presented above A 2 Example of the prepsource script Let us now describe a typical interactive prepsource session ie with ENVIRONMENT INTERACTIVE Below a succession of steps are presented in which a user has used the full possibilities of prepsource Lines preceded by number are output from prepsource whereas other lines correspond to the user s answers 1 gt please enter a name for your simulation directory pwd FULL SCREEN VISUALIZATION OF THE CONTROL PARAMETERS m the user can again proceed to some modifications gt gt user library specified 3 gt Carriage Return alone to valid 3 gt something else and CR to stop CR 4 users come fischer mnh master0 make mnh prov does not exist default is searche 4 gt d 4 gt enter the root directory from which source files will be extracted 4 gt default 4 ist 4 2nd users come fischer mnh masterO 4 3rd mesonh sources 4 gt HOME sources modif 5 gt please enter arguments for nh takeout file name version name VID mod dein f90 MASTERO 651 1 retrieved 7 gt do you want
55. fixing the free parameters for the dynamics and the physics of the Meso NH model This allows the user to know a large part of the history of this file For the namelists or variables ommited in the EXSEG n nam file the values are set to the default ones see the tables in ch 4 If the file is the result of the initialization programs PREP IDEAL CASE PREP REAL CASE or SPAWNING the values of the namelists variables are the ones of the descriptive part of the input file of the program if it does exist Otherwise the values are set to the default ones except for these that can be initialized during the initialization program e g CINIFILE or LUSERV Note that a physiographic file does not have a descriptive part binary part All the writings and readings of this type of files are done through LFI routines general subroutine to read and write a Meso NH file is given in the Meso NH library it provides a file including the fields of the previous record list This Fortran library provides a way to tackle direct access binary files and thus a very quick access to the data stored in this file in any order It should be noted that supplementary fields can be added to these basic informations which have been obtained at the same instant In order to be easily drawn by the Meso NH graphic package the commentar field must be filled according to the following rules e the length of the character string is equal to 100 e the type of
56. for analysis and first guess base time for forecast STEP variable used for forecast files it corresponds to the length of forecast since TIME Both next variables will allow the user to execute a single extractarpege for operational archive for several consecutive dates beginning from DATESTIME for TYPE AN or from DATESTIMESSTEP for TYPE FC NBLOOP number of more dates after SDATESTIME or DATESTIMESSTEP 0 will do an ordinary extractarpege with one single date LOOPSTEP interval between each date extracted HH The five next variables are needed for special archive different from operational one but given in ARPEGE or ALADIN format e g some experiment reanalyses INLOGIN user on the M t o France storage file archiv machine where the input file is stored INDIR directory of the input ARPEGE or ALADIN file starting at HOME for archiv machine home starting at HOME for super computer storage work starting at workdir for super computer storage INFILE name if the input ARPEGE or ALADIN file LMAX number of vertical levels present in the input file SOIL name of the soil scheme used for the input file generation ISBA OISB NONE or other OUTFILE name of the output GRIB file When several dates are requested the number of output file is NBLOOP 1 The name is set to MODEL TYPE DATE TIME if TYPE AN MODEL TYPE DATE TIME STEP if TYPE FC OUTHOST output file will be put onto loca
57. if you want to initialize the surface prog nostic fields from a input file You can also define more precisely the surface fields by using the namelists for each scheme 5 10 MISCELLANEOUS INTERACTIVE PROCEDURES 175 5 10 Miscellaneous interactive procedures set of procedures runing interactively are available with the MesoNH environment 5 10 1 Content of MESO NH file fmmore This procedures lists the fields stored in a FM file MESONH or PGD file gives the projection and domain definition the number of points in each direction the horizontal grid meshes the entire vertical grid and the time informations of the file The descriptive part of the FM file is also printed see an example in 8 2 1 This procedure is used as fmmore MESONH file name 5 10 2 Definition of the vertical grid vergrid In order to help the user to define ist vertical grid with the FUNCTN type this procedure computes the new grid as done in PREP IDEAL CASE or PREP_REAL_CASE It uses a namelist file containing only the namelist NAM VER GRID containing the vari ables LTHINSHELL NKMAX YZGRID TYPE ZDZGRD ZDZTOP ZZMAX STRGRD ZSTR GRD ZSTRTOP described above in PREP REAL CASE presentation This procedure is used as vergrid namelist file name 5 10 3 Conversion from spherical coordinates to conformal coordinates lat lon2xy This procedures computes the position of points given with their latitude and longitude onto the confo
58. informations written in the prepsourcerc appears at the screen You do not need to change anything Validate the 2 pages with a blanck character after each page The second question concerns the localization of the RCS libraries DIRSRC The screen is as follows enter the root directory from which source files will be extracted default 18 2nd users mesonh stein SOURCE 3rd mesonh sources 4th mesonh bugfix masdev4 1 The 2nd directory represents the desired directory DIRSRC and you only enter a carriage return Then the procedure asks you the name of the RCS library from where you want to take out the Fortran source please enter arguments for nh takeout file name version name VID now you enter one RCS library name cfl f90 EXAMPLE the answer refers to the RCS library cfl f90 v and to the version defined by the EXAMPLE name The procedure gives you the number of the desired version 1 1 Retrieved 1 1 436 lines do you want another file y or n default y and asks you whether you want to compile another Fortran source or not Enter y 210 CHAPTER 7 MODIFY THE FORTRAN SOURCES You come back at the question where the directory for source location is prompted answer by a carriage return and then by modeln f90 EXAMPLE to take out the second Fortran source Then you go out of the infinite loop by answering n to the question The presource procedure asks you enter the global parameters fo
59. is perform by the standard UNIX file manager RCS e procedures are written in standard UNIX Korn Shell e The generalized use of dynamic memory allocation avoids repeated compilation of the source and the free parameters are set by the user through namelist files sequence of elementary actions are to be done to realize a complete numerical experiment as illustrated by Figure 1 1 The first part is used to modify the Fortran sources of the MESONH library or to add new Fortran sources in order to adapt the model to the user s wishes Nevertheless because Meso NH is able to perform a lot of actions without any modifications of the Fortran sources we will end the presentation of the book with this first part when the new user will know how to manipulate Meso NH as it stands The second part corresponds to the generation of the 8 CHAPTER 1 INTRODUCTION MODIFY A FORTRAN SUBROUTINE PR ASE PR CASE PREPARATION OF A SIMULATION SIMULATION 3 INTEGRATION e MODEL N DIAGNOSTICS COMPUTATION Figure 1 1 General algorithm for a complete numerical experiment initial MESONH files which contain the values of the prognostic variables of the model This is realized by 3 differents programs which correspond to the preparation either of an ideal case study idealized atmosphere and orography either of a real case study interpolation from a Larger Scale operationnal model to the finer mesh
60. is why the following list starts at level 2 level 2 ALTITUDE variable TEMPERATURE variable MOIST variable additional cloud variable s the mass level 1 is at ground level ALTITUDE variable TEMPERATURE variable MOIST variable additional cloud variable s uppermost mass level ALTITUDE variable TEMPERATURE variable MOIST variable additional cloud variable s Example of free part of PRE IDEA 1 nam RSOU 1990 10 3 72000 gt STANDARD 200 100240 287 5 276 2 85000 20 10 70000 30 10 3 90000 280 275 60000 271 269 3 3 8 Constant moist Brunt Vaisala case keyword CSTN Data of the vertical profile are written in the free format part of PRE IDEA1 nam file in the following order 40 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE YEAR integer example 1994 MONTH integer example 4 DAY integer example 22 TIME real in seconds example 36000 for 10 h NUMBER of LEVELS integer VIRTUAL POTENTIAL TEMPERATURE at GROUND LEVEL i e at the first level realin Kelvin PRESSURE at GROUND LEVEL i e at the first level real in Pascal HEIGHT at all levels the first level is the ground level ZONAL WIND COMPONENT at all levels the first level is the ground level MERIDIAN WIND COMPONENT at all levels the first level is the ground level RELATIVE HUMIDITY at all levels the first level is the ground level MOIST BRUNT VAISALA FREQUENCY a
61. n from 4 2 11 e Thirdly namelists relative to the surface model from 4 2 21 e Fourthly namelists relative to chemistry from 4 2 22 e Fifthly namelists relative to on line diagnostics balloons aircrafts profilers stations series from 4 2 25 e And endly namelists relative to budgets from 4 2 29 and LES budgets from 4 2 43 42 THE INPUT EXSEGS N NAM FILE 53 4 2 1 Namelist NAM BLANK available variables Foran type default value XDUMMYS real NDUMMY8 integer LDUMMYS logical array real 0 0 TRUE CDUMMYS 80 characters D 20 0 array integer 20 0 array logical 20 TRUE array 80 characters 20 Eight dummy real integer logical and character 80 variables and arrays of dummy real integer logical and character 80 for test and debugging purposes are defined and passed through the namelist read operations None of the MesoNH routines uses any of those variables When a developper choses to introduce temporarily a parameter to some subroutine he has to introduce a USE MODD BLANK statement into that subroutine Then he can use any of the variables defined here and change them easily via the namelist input 4 2 2 Namelist NAM CONF global configuration parameters It contains the model configuration parameters common to all the models are included in the module MODD CONF Fortran type default value CCONF 5 characters LFL
62. now ready to start to work Copy in this directory all the files present in MESONH procedures ending by rc in SIMUL In this example we will need to modify prepmodelrc tosuprc with any editor to obtain the following files FILE prepmodelrc 3 5 AN ILLUSTRATIVE EXAMPLE OF PREP IDEAL CASE 45 bin sh default input variables for prepmodel Asencio 21 09 94 d NEW Improvements in prepmodelrc are added at the beginning of the file read them d comfort variables S0UTSCRIPT instead of outprepmodel outprepideal outprep_pgd outprepreal outprepspawn outprepdiag SLEEPSECONDS instead of 10 EDITOR vi specific variables Control tools commands in prepmodel global var yes or no default variable TOOLSCONTROL no tori hpce tora fuji until next monday 1x____ 32bit a must have OUTDEST Automatic submit by tosupc for part one tosupcrc with TIME MEM NBP TPN has to be OK LSOUMISAUTO T options for treatment of all FM files and post treatment 4 lfiz 21112 conv2dia fm unlfiz all fmmore OUTFILE TOOLS SUBMIT NEXTJOBS O or filenames to submit super calculator at the end of this job tosupcrc must be present on the same directory job one or several scripts between commas which will be submitted one by one by tosupc the pathnames are relative to the current SIMUL directory or absolute dk dt dk dt dt dt OH SUBMIT NEXTJO
63. of constant altitude levels where the local quantities are computed Not used per default NSPECTRA LEVELS list of model levels where the non local quantities are computed Any number is allowed but too many will be costly in CPU time and memory XSPECTRA HEIGHTS list of constant altitude levels where the non local quantities are computed Any number is allowed but too many will be costly in CPU time and memory NLES TEMP SERIE I list of the I coordinates model index of the points where temporal series are extracted Not yet implemented NLES TEMP SERIE J list of the J coordinates model index of the points where tem poral series are extracted Not yet implemented NLES TEMP SERIE Z list of the altitudes of the points where temporal series ex tracted Not yet implemented CLES NORM TYPE type of normalization of the fluxes and variances NONE no normalization is computed however the quantities necessary to perform these are computed and stored in the file CONV convective normalisation using Qo ws h lt w r gt EKMA Ekman normalization using u and LEkman MOBU Monin Obukhov normalization using Lo us Qo lt w r surf CBL HEIGHT DEF definition of the Boundary Layer height h KE 7 test on total kinetic energy E h e h 0 054 S E z e z dz WTV test on lt w 0 0 gt height h where this flux is most negative DTH
64. the model TRUE The vertical relaxation is applied FALSE The vertical relaxation is not applied NRIMX number of points included in the lateral relaxation area in the x direction NRIMY number of points included in the lateral relaxation area in the y direction XRIMKMAX maximum value in s of the relaxation coefficient for the lateral relax ation area This value is appllied on all the outermost verticals of the domain Caution this value is also used to relaxe the normal wind for open lbc condition This relaxation exists in the Carpenter equation even if LHO RELAX UVW TH F X TADIFF characteristic time e folding time for the numerical diffusion of fourth order seconds Associated to LNUMDIFF in NAM DYN 42 THE INPUT EXSEGS N NAM FILE 67 4 2 14 Namelist LBCn boundary conditions of model n Fortran type default value CLBCX array 2 characters 2 CYCL CLBCY array 2 characters 2 CYCL XCPHASE real 20 It contains the parameters needed to specify the lateral boundary conditions for the model n They are included in the declarative module MODD LBCn e CLBCX represent the type of lateral boundary condition at the left and right boundaries along x CLBCX 1 and CLBCX 2 respectively The possible values are CYCU for cyclic boundary conditions this case CLBCX 1 CLBCX 2 CYCL OPEN for open boundary condition Sommerfeld equation for the norma
65. the HATMFILE or if the ones of the HATMFILE are not to be used only if HATMFILETYPE is GRIBEX The grids must be the same as the ones of the output file CINIFILE e HCHEMFILETYPE type of the chemical file GRIBEX MESONH e CINIFILE name of the MESO NH output FM file used as initial or coupling file in a MESO NH simulation 2 Namelist NAM REAL CONF contains configuration variables CEQNSYS character LEN 3 if HATMFILETYPE GRIBEX DUR if HATMFILETYPE MESONH CEQNSYS value used in input MESONH file CPRESOPT character LEN 5 CREST NVERB integer 1 LSHIFT logical if HATMFILETYPE GRIBEX TRUE if HATMFILETYPE MESONH FALSE CEQNSYS EQuatioN SYStem LHE Lipps HEmler 1982 Modified Anelastic Equations following DURran 1990 derivations CPRESOPT option for pressure solver RICHA CGRAD CRESI e NVERB verbosity level error diagnostics are computed if NVERB gt 4 e LSHIFT flag to shift altitudes in boundary layer Namelist NAM VER GRID contains vertical grid definition The use of the THINSHELL approximation is specified in this namelist There are five ways to compute the vertical grid the three first ones are as in PREP IDEAL CASE a constant grid mesh only the number of levels NKMAX and the grid mesh sizes ZDZ GRD and ZDZTOP are used These must be equal The type of grid YZGRID_TYPE is set to PUNCTN
66. the model n They are included in the declarative module MODD_PARAM_RADn e XDTRAD Interval of time in seconds between two full radiation computations the radiative tendency is computed for all the verticals of the simulation domain This is done in order to save CPU time because the radiation scheme is very expensive but the radiative tendency does not evolve too much in some cases during periods greater than the model timestep XTSTEP In this case the radiation timestep is increased to XDTRAD e XDTRAD CLONLY Interval of time in seconds between two radiation computations for the cloudy columns only This is based on the same principle as the intermittent full radiation call the cloudy column radiative tendency may in some cases evolve faster than the dry ones but still slower than the timestep XTSTEP In this case the cloudy radiation timestep is increased from XDTRAD to XDTRAD CLONLY Of course when all and part of the radiative tendencies must be refreshed during the same MESONH timestep only the full radiation call is performed e CLW choice of LW code RRTM RAPID RADIATIVE TRANSFER MODEL MORC MORCRETTE model e CAER type of aerosol distribution SURF deduced from cover data 72 CHAPTER 4 PERFORM A MESONH SIMULATION computed from Tegen et al 1997 mensual climatology horizontal resolu tion is 4 for latitude and 5 for longitude computed from
67. the screen the control parameters the user moves from one field to another using the arrow keys once a field is changed it is validated by a CR to leave the page and to go on the space key has to be pressed Step 2 is equivalent to prepsource step 3 prepmodel is simply concerned about the user not having indicated any user library name but 0 This may not be an error if 0 is retained this simply means that all the compiled object code will be extracted from the master binary library Of course for the general use and task during development or diagnostic investigation with the Meso NH code the user will indicate his own binary library Step 4 allows the user to verify and modify the command lines used to get the input files a file containing this code is edited because GETCONTROL yes This could prove helpful if special files have to be got In cases where ENVIRONMENT SILENTINTERACTIVE or BATCH mesonh will not ask anything to the user nor read in any file a file inprepmodel does not exist In SILENTIN TERACTIVE prepmodel will still print output information whereas it is dumb in BATCH Finally let us emphasize some specific points First if the control variable MAINPROG is set to MODEL mesonh will try to read some information the name of the Meso NH input data files the number of models in the namelist formatted file EXSEGI1 nam and EXSEG2 nam etc in multitask Thus this file has absolutly to be present on the
68. the supplementary field must be specified SS ommend Dsm XX Zvameme UNIT 2D scar XXYsmmame UNIT 3D vector VX xvarname VY yvarname VZ zvarname UNIT 2D scalar VX xvarname VY yvarname VZ zvarname UNIT VX xvarname VY yvarname UNIT IDswhr ON Content of a synchronous file with fmmore As explained in section 5 10 1 this procedure gives information about the MesoNH file First it lists the records present in the binary part then the values of some parameters present in the binary part e g the dimensions of the fields and finally the descriptive part of the file This set of informations can be obtained for any Meso NH file we give here an example for a synchronous file You may use the procedure fmmore by the following order fmmore FILENAME 81 1 256 125696 LFILAF Catalogue de l Unite Logique LFI 12 dans l ordre PHYSIQUE sequentiel des articles 3 article s physique s de gestion 512 mots chacun occupant donc 1536 mots detail Article documentaire de la position 1a 512 216 CHAPTER 8 THE MESONH FILES 1 paire s d articles d index prereserves de la position 513 a 1536 pete pas de paire d articles d index inutilises ni excedentaires sss 1 eme article de donnees MY NAME 50 mots position 1537 a 1586 2 eme article de donnees DAD NAME s 50 mots position 1587 a 1636 3 eme ar
69. to give the localization of vertical profile string of 6 characters DJGRID for i j point on index space XYHATM for x y coordinates on conformal plane or cartesian plane LATLON for latitude longitude on spherical earth e XLATLOC Latitude in degrees of the vertical profile localization used in case CTYPE LOC LATLON e XLONLOC Longitude in degrees of the vertical profile localization used in case CTYPELOC LATLON e XXHATLOC position in meters x of the vertical profile localization used in cases CTYPELOC XYHATM 34 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE e XYHATLOC position in meters y of the vertical profile localization used in cases CTYPELOC XYHATM e NILOC position i of the vertical profile localization used in cases CTYPELOC IJGRID If you use a 1D model then NILOC is reset to 2 by the program e NJLOC position of the vertical profile localization used in cases CTYPELOC IJGRID If you use a 1D or 2D model then NJLOC is reset to 2 by the program 3 2 18 Namelists of the externalized surface The further definition of the surface parameters is not done by MESONH itself but by the externalized surface included in it So you are invited to refer to the documentation of the surface Two cases are encountered 1 You do not have any input PGD file or you do not want to use the surface fields in cluded in it LREA
70. variables which must be initialized before running the procedures or as input parameters for the procedures themselves 2 1 2 The environment variables The environment variables are necessary for the UNIX procedures to recover the right paths where the informations must be taken out The environment variables have to be set up in the profile of the user on the local host and are listed below 1 ENVIRONMENT indicates how the MESONH scripts will be run There are 4 possible values e INTERACTIVE the scripts which prepare the job that will be executed later on the remote host are called during an interactive session on the local host Full screen 14 CHAPTER 2 THE MESONH PROCEDURES EXCEPT PREPSOURCE LOCAL HOST REMOTE HOST LOCAL HOST MOTE HOST SIMUL prepsourcerc prepmodelrc TEMP executable model code PRE_IDEA1 nam bibmaster a ABS MAINPROG exe EXSEG1 nam DIAG1 nam FILE A FILE B Figure 2 2 different locations of the informations initialization is allowed and some questions may be asked to the user INPUT files keyboard and OUTPUT screen e SILENTINTERACTIVE the scripts are called in an interactive session but no full screen initialization and no questions are asked INPUT files and OUTPUT Screen e BATCH the scripts are called in batch mode INPUT files and OUTPUT file e CRON the scripts are
71. we therefore decide to write complete namelists filled by the default parameters when these variables were not initialized 44 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE In conclusion not all the informations present in the descriptive part of the initial MESONH file are significant The specification of the segment to perform with this MESONH file used as starting point will be realized in the EXSEGI nam file described in chapter 4 3 5 illustrative example of PREP IDEAL CASE This section presents a complete example of a preparation of initial conditions for an academic study it explains how to build on the localhost the UNIX job which will execute the fortran program on the remote_host selected case is the following e 2D mountain e moist one layer atmosphere The first step is the control of the initialization for the control variables Enter the following commands on the localhost when you are logged in echo SIMUL this is the working directory If it is not the right name change it by SIMUL myworkdir export SIMUL now go in this directory cd SIMUL In one configuration file profile bash profile set the other environmental variables e PATH must contain MESONH procedures or enter the following command PATH PATH MESONH UPDATE PROCEDURES MESONH procedures e MANPATH must contain MESONH procedures or change it as for PATH e EDITOR usr bin vi You are
72. within the PREP REAL CASE program i The hurricane filtering is applied on four input atmospheric Grib fields HATM FILETYPE GRIBEX when they in the horizontal grid of the PGD file and in the vertical grid of the Grib file The input atmospheric Grib fields filtered are the two horizontal components of wind the absolute temperature and the surface pressure re duced to ground level Each field is decomposed into three parts first the BASic part is computed by the low pass Barnes filter then the hurricane symmetric disturbance is computed from the remainder disturbance part initial fields are then remplaced by their ENVironmental part total field minus hurricane disturbance part ii The vortex bogussing consists on a symmetric vortex added to the input atmo spheric MesoNH fields HATMFILETYPE MESONH The tangential wind is computed from an analytical formulation Holland 1980 Then the balanced mass field is deduced from the thermal wind relation The bogus of the two horizontal components of wind and the potential temperature is added to the initial filtered fields Thus two PREP REAL CASE jobs to be performed In a mono model configuration the first job allows to remove analysed hurricane from the input Grib fields filtered and interpolated fields are written in a MesoNH file It is used as input file for the second PREP_REAL_CASE job during which the analytical vortex is added 164 CHAP
73. work unless you correctly fill TIME MEM TPN NBP time in seconds for J2 amp J3 jobs TIME memory ex 2000 2Gb 128Mb tori less than 128Gb per node be careful reduce the memory for multi tasks jobs MEM tasks per node only for prep ideal case run or diag number of used CPUs from 1 to 8 The most important for parallel execution TPN 1 mono or multi node on NEC 1 to 4 and on IBM NBP 1 THEEEBHHHHHEHHHHHHEHHE tosupc will not work unless you correctly fill TIME MEM TPN NBP Sending a mail abort beg and or at the end job begend beg end no default is abort no is no mail at all MAIL global variables used in tosupc DEBUGSCRIPT 0N OFF ENVIRONMENT SILENTINTERACTIVE SILENTINTERACTIVE BATCH INTERACTIVE All the necessary informations for the UNIX to be correctly executed on the remote host will be read during this preparative job on the different rc files on the local host Now you may create a file which contains the following namelists FILE PRE IDEA1 nam amp NAM DIMn PRE 128 NJMAX 1 VER GRID NKMAX 32 YZGRID TYPE FUNCTN ZDZGRD 500 ZDZTOP 500 ZZMAX STRGRD 1000 ZSTRGRD 0 ZSTRTOP 0 CONFn LUSERV TRUE NSV USER O GRID PRE XLATO 48 25 XLONO O XRPK O XBETA O XLONORI 48 25 XLATORI O CONF PRE LCARTESIAN TRUE LBOUSS FALSE CIDEAL CSTN CZS BELL LPERTURB FALSE
74. you have to enter the name of the super computer in which the job will be executed In SILENTINTERACTIVE or BATCH tosupc will ask nothing to the user but simply use the information in tosupcrc to work Also in both modes tosupc will not ask for the NQS parameters and you thus have to initialize them correctly in the tosupcrc file an example of such a file is stored on MESONH procedures Index C CAER in namelist NAM_PARAM_RADn 69 CBUTYPE in namelist NAM BUDGET 81 CCH_TDISCRETIZATION in namelist NAM CH MNHCn 74 CCH_TS1D_COMMENT in namelist NAM CH MNHCn 75 CCH TSID FILENAME in namelist NAM CH MNHCn 75 CCH TUV CLOUDS in namelist NAM CH MNHCn 75 CCH TUV LOOKUP in namelist NAM CH MNHCn 75 CCH VEC METHOD in namelist NAM CH MNHCn 75 CCHEM_INPUT_FILE in namelist NAM CH MNHCn PRE 22 in namelist NAM CH MNHCn 74 CCLOUD in namelist NAM_PARAMn 66 CCONF in namelist _ 51 CCPLFILE in namelist NAM LUNITn 65 CDADATMFILE in namelist NAM HURR CONF 161 CDADBOGFILE in namelist NAM HURR CONF 161 CDCONV in namelist NAM PARAMn 67 CEFRADI in namelist NAM_PARAM_RADn 69 CEFRADL in namelist NAM_PARAM_RADn 69 CEQNSYS in namelist NAM_CONF_PRE 23 in namelist NAM CONF 51 in namelist NAM REAL CONF 157 CEXP in namelist CFILTERING in namelist CFUNU in namelist CFUNV in namelist CIDEAL in namelist CINIFILE in namelist in namelist in namelist CINIT LG in namelist CLBCX in n
75. 0 for aircraft 82 CHAPTER 4 PERFORM A MESONH SIMULATION 4 2 26 Profilers and stations In order to compare the model outputs to observations and measurements it can be interesting to store the simulated data for a given profiler or station profiler is located at a given location with its latitude and longitude whereas a station is located with its latitude longitude and altitude Pronostic fields are recorded at a time frequency prescribed zonal and meridien wind from U and V components vertical velocity potential temperature pression mixing ratios tke radiative surface temperature If surface diagnostics are asked see surface namelists surface variables 10m wind 2m temperature and humidity surface fluxes are also stored records are in the diachronic file 000 The specification of the characteristics of profilers and stations are not given in a namelist but directly in Fortran routines e ini profilern f90 for profilers e ini_stationn f90 for stations 4 2 27 Namelist NAM SERIES temporal series in diagnostic file Fortran type default value LSERIES boolean LMASKLANDSEA boolean LWMINMAX boolean e LSERIES switch to write temporal series in the diachronic file 000 of each model evolution of horizontaly and verticaly averaged fields t evolution of horizontaly averaged vertical profils z t evolution of y horizontaly averaged fields at one level or verticaly averaged between 2 l
76. 1 ETEP 232 1 OLTP 233 1 OLIP 234 1 ISOP 235 1 APIP 236 1 LIMP 237 1 PHO 238 1 ADDT 239 1 ADDX 240 1 ADDC 241 1 TOLP 242 1 XYLP 243 1 CSLP 244 1 ACO3 245 1 TCO3 246 1 KETP 247 1 OLNN 248 1 OLND 249 1 X02 END CONVMOC2GRIB keyword a comment line CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES The format used in MOCI nam is fixed that is the number of chemical species in the output Grib file for each output species Grib code number fraction of input species input species name e g 242 1 NO 1 NO_2 2 N 20 5 a keyword for the end 5 7 VERTICAL INTERPOLATIONS PREP_REAL_CASE 159 5 7 Vertical interpolations PREP REAL CASE PREP REAL CASE performs the vertical interpolations from one orography to another The main hypothesis is that hydrostatism is verified Therefore if the input file is MESO NH file there is small loss of information It is possible to use this program to modify the vertical grid without changing the orography What s going in and out e Input the prepmodelrc file a file containing the atmospheric 3D and surface 2D variable fields hereafter called atmospheric file it can be either an GRIB file obtained from extractecmwf or extractarpege or a MESO NH file obtained with SPAWNING for example a physiographic data file it can also be a complete MESO NH file an optional file containing the chemical species here after called chemi
77. 38 Namelist NAM BU RRI budget for non precipitating 92 4 2 39 Namelist NAM BU RRS budget for snow 93 4 2 0 Namelist NAM BU budget for graupel 93 4 24 Namelist NAM BU budget for 94 4 2 42 Namelist NAM BU RSV budget for a Scalar Variable 94 4 2 43 LES budgets Namelist NAM LES a 94 4 2 4 LES budgets Namelist NAM PDF 98 LES diaenostics EURO REOR el eed 99 4 3 1 TMOTALIONS n EUR Ge RES 3 We Pe eek a RUE A 99 43 2 WMhatasaveulable eumd XI arsi 100 43 3 LES averaged fields LLES MEAN TRUE 100 ASA LES pdt LEBS PDE DRUBE kt om ee Ren 3s 101 4 3 5 LES averaged fields LLES RESOLVED TRUE 101 4 3 6 LES averaged fields LLESSUBGRID TRUE 105 43 7 LES averaged fields LLESUPDRAFT TRUE 105 CONTENTS 4 3 8 LES averaged fields LLES DOWNDRAFT TRUE 107 4 3 9 LES averaged surface fields 108 4 3 10 LES 2 points correlations 109 ASV EES Spectra ca cs sec SP eg a eS 110 4 4 Budget of resolved subgrid turbulent quantities 111 441 Budget of total turbulent kinetic energy 111 4
78. 4 2 Budget of total liquid temperature flux 2 114 443 Budget of total liquid temperature variance 116 4 4 4 Budget of total water flux 118 4 4 5 Budget of liquid temperature total water covariance 120 4 4 Budget of total water variance 122 447 Budget of total scalar 2 124 4 4 8 Budget of total scalar variance 126 4 5 An illustrative example of a MESONH simulation 128 Initialization of MESO NH for real cases 135 5 1 Overview of the initialization sequences 135 5 1 1 Initialization for a one model run from operationnel models 135 5 1 2 Initialization for a one model run from a MESO NH file 138 5 1 3 Initialization for a nesting run 138 5 2 Creation of MESO NH physiographic data PREP PGD 144 5 24 Execution yon po RUE Pine MUI et 144 5 2 2 Presentation of input files 144 5 2 3 The input PREP_PGDI1 nam file 145 5 2 4 Examples of 1 file 145 5 3 Modification of PGD files for grid nesting PREP NEST PGD 148 5 4 Zoom of a PGD file ZOOM_PGD 150 5 5 Extraction of ECMWF files extractecmwf 152 5 6 Extraction of M t o Franc
79. 5 0 00190 0 00300 0 0 0 0 3 3 5 Discretized orography keyword ZSDATA Only the orography corresponding to the computational domain must be provided in free format For 3D orography data are read like if it was a map the first line is the Northern border and the first data is the North West corner with one line per Y axis increment Example of free part of IDEA1 nam ZSDATA 30 30 35 50 30 30 30 59 5 133 3 100 2 136 7 100 35 89 5 183 3 200 2 299 7 170 5 50 112 5 193 0 210 2 206 7 120 40 82 5 153 0 180 5 156 7 100 3 3 4 The output MESONH file The descriptive part file DESFM of the MESONH file is built with the default values associ ated with all the degrees of the model simulations and with some informations present in the PRE_IDEA1 nam file name of the file activations of the different moist variables There fore most of the informations present in the descriptive part DESFM for instance type of the turbulence amplitude of the Rayleigh damping are not really related to the description of the generation of the LFI file This is due to the way of building this initial file not from a MESONH simulation but from external informations given in PRE_IDEA1 nam A better way of filling this descriptive part would have been to generate empty namelists for this kind of informations in the LFI file but in Fortran 90 it is still impossible to write empty namelists they have to include at least one element
80. 7 ATSR 1to3 1to3 MHS 1 05 1 05 IASI 1 to 8461 1 to 8461 AMSR 1 to 14 1to7 MVIRI 1to2 1to2 SEVIRI 4 to 11 1to8 GOES Imager 1to4 1to4 GOES Sounder 1 to 18 1 to 18 GMS MTSAT imager 1to4 1 to 4 FY2 VISSR 1 to 2 1 to 2 FY1 MVISR 1 to 3 1 0 Cris TBD TBD CMISS TBD TBD VIIRS TBD TBD WINDSAT 1 to 10 1to5 1 00 1 b2 2 b2 b2 b2 b2 KP ER EFE 2 2 1 DOK WD Table 6 2 Instruments supported by RTTOV 8 7 as at 17 Nov 2005 Sensors in italics are not yet supported by RT TOV 8 7 but soon will be taken from RTTOV 8 7 Users guide Table 3 page 5 Handling satellite data into the Meso NH world Satellite data can be obtained from many different archive centers and with diffferent formats e One of them is the French archive center called SATMOS Service d Archivage et de Traitement M t orologique des Observations Spatiales see http www satmos meteo fr SATMOS can provide observations from NOAA GOES GMS and METEOSAT satellites See the quick look pages and send an email to satmosQmeteo fr to order satellite images do not forget to specify the channels the TARCYL format the interest area the date The program readtarcyl below gives an example in reading the TARCYL format 6 3 VARIABLES AVAILABLE IN THE OUTPUT DIACHRONIC FILE 193 e An alternative way to METEOSAT data is through EUMETSAT see http www eumetsat int Observation from NOA
81. 9 15 or 21 STEP can be 03 or 09 the length of forecast since TIME for forecast file see MARS documentation for allowed values 6 hourly to TIME 120 then 12 hourly to 240 for dates before 14 11 90 3 hourly to TIME 12 then 6 hourly to TIME 120 and 12 hourly to TIME 240 for dates since 15 11 90 e AUTO automatic definition of grid and area see below y n Automatic let you get the maximum resolution available in the archive without interpola tion e AREA output area definition MARS parameter N lat W long S lat E long or G for global or E for Europe 73 5 27 0 33 0 45 0 Default for AUTO y is global e GRID output grid mesh definition MARS parameter 2 values give E W and 5 incre ment latitude longitude grid minimum mesh is 5 5 or a single positive integer specifying the number of line of latitude between the Pole and the Equator in Gaussian output For AUTO y the maximum resolution of Gaussian grid is automatically set according to the date Both next variables will allow the user to execute a single extractecmwf for several consecu tive dates beginning from DATE TIME for TYPE AN AN or from DATE TIME STEP for TYPE FC FC e NBLOOP number of more dates after SDATESTIME DATESTIMESSTEP 0 will do an ordinary extractecmwf with one single date 154 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES LOOPSTEP interval between each date extracted HH Both next v
82. 96 LSLEVE logical FALSE XLENI real 7500 XLEN2 real 2500 e LTHINSHELL switch for the thinshell approximation logical e NKMAX number of points in z direction of the required physical domain The total size of the array written in initial file will be VK MAX 2JPVEXT JPVEXT is fixed to 1 for the present version of Meso NH e YZGRID_TYPE type of vertical grid definition FUNCTWN the levels are calculated by the program according to the namelist variables MANUAT the levels are written in the free formatted part after the namelist 5 7 VERTICAL INTERPOLATIONS PREP_REAL_CASE 163 SAMEGRE the levels are the same as those in the input file Only available when atmospheric input file is MESONH file e ZDZGRD mesh length in z direction near the ground e ZDZTOP mesh length in z direction near the top of the model e ZZMAX STRGRD Altitude separating the two constant stretching layers e ZSTRGRD Constant imposed stretching in 96 in the lower layer below ZZMAX_STRGRD e ZSTRTOP Constant imposed stretching in 96 in the upper layer above ZZMAX STRGRD e LSLEVE flag for Sleve vertical coordinate e XLENI decay scale for smooth topography in meters e XLEN 2 decay scale for smale scale topography deviation in meters 4 namelist NAM HURR defines hurricane filtering and vortex bogussing Each step hurricane filtering and vortex bogussing is separately invoked
83. A satellites and many others can be also obtained CLASS Comprehensive Large Array data Stewardship System see http www class noaa gov The ATOVS and AVHRR Pre processing Package see http www metoffice com research interproj nwpsaf aapp processes data from instruments on board the NOAA polar orbiting satellites namely HIRS AVHRR AMSU and MHS on the current generation of satellites e Finally the satellite data can be projeted onto a Meso NH grid by using the command obs2mnh from the libtools package Program readtarcyl PROGRAM readtarcyl acce co rrr DR ERR Read METEOSAT MSG GOES INFRARED data in TARCYL format and write outputs to be readen by the obs2mesonh tool AUTHOR u L J P Chaboureau L A MODIFICATIONS USE 90 UNIX IMPLICIT NONE INTEGER II IUNIT IO INARG IUOUT IUDIR INTEGER JI JJ NBYTE IXSIZE IYSIZE INIL IVAL REAL ZLON ZLAT ZVAL REAL ZLATMIN ZLATMAX ZLONMIN ZLONMAX ZVALMIN ZVALMAX CHARACTER LEN 80 YFNAM YRAC YLINE YMEM YDIR YOUT CHARACTER LEN 21 YLBLCH CHARACTER LEN 12 YSAT CHARACTER LEN 2 YTMP2 CHARACTER LEN 1 IUNIT 12 IUDIR 40 IUOUT 41 YMEM def YDIR satobs dir YOUT satobs asc INARG IARGCO IF INARG 0 THEN PRINT Usage readtarcyl tarcyl file STOP END IF CALL GETARG 1 YRAC YFNAM TRIM YRAC def 194 CHAPTER 6 COMPUTE DIAGNOST
84. ACcumulated INstantaneous Precipitation Rates explicit Snow mm mm h ACPRG INPRG 2D ACcumulated INstantaneous Precipitation Rates 52 Graupel mm mm h ACPRH INPRH 2D ACcumulated INstantaneous Precipitation Rates a Hail mm mm h ACPRT INPRT 2D ACcumulated INstantaneous Precipitation Rates x all explicit species mm mm h PACCONV PRCONV 2D ACcumulated instantaneous Precipitation Rates 5 total CONVective mm mm h PRSCONV 2D instantaneous Precipitation Rate mm h Solid CONVective ACTOPR 2D Accumulated Total Precinitation mm PR 6 3 VARIABLES AVAILABLE IN THE OUTPUT DIACHRONIC FILE 187 Variables dim meaning unit in DIAG1 nam CAPEMAX 2D maximum of CAPE3D J kg NCAPE 0 1 2 CINMAX 2D value of CIN3D corresponding to CAPEMAX J kg i CAPE3D CIN3D DCAPE3D 3D J kg NCAPE 1 2 VKE 3D Vertical Kinetic Energy from explicit vertical motion J kg NCAPE 2 BV BVE 3D Brunt Vaissala frequencies s LBV FR RARE 3D Radar Reflectivity dBZ LRADAR VDOP 3D radar Doppler fall speed m s ZDR 3D radar Differential Reflectivity dBZ KDP 3D radar Differential Phase shift degree km HBLTOP 2D Height of boundary layer top m KBLTOP 2D Index of boundary layer top FREE ATM GR 2D Gradient of free atmosphere above BL top K m THV_FREE 3D Thetav above BL top K Surface parameters 25 2D orography m ZSMT
85. AM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU RRI 89 RRR 88 RRS 90 RRV 86 RSV 91 RTH 85 RTKE 86 RU 83 RV 84 RW 84 RRR 88 RRV 86 RTH 85 RRC 87 RRG 90 RRS 90 RTH 85 NAM DYNn 64 NAM DYNn 64 RRC 87 RRG 90 RRH 91 RRI 89 RRR 88 NSEDIRS in namelist NSFRRG in namelist NSFRRR in namelist NSFRTH in namelist NSLEVE in namelist NSV_USER in namelist NTPTKE in namelist NTRTKE in namelist NVERB in namelist in namelist in namelist NVTURBRC in namelist NVTURBRI in namelist NVTURBRV in namelist NVTURBSV in namelist NVTURBTH in namelist NVTURBU in namelist NVTURBV in namelist NVTURBW in namelist NWETGRC in namelist NWETGRG in namelist NWETGRH in namelist NWETGRI in namelist NWETGRR in namelist NWETGRS in namelist 239 NAM BU RRS 90 NAM BU RRS 90 NAM BU RRR 88 NAM BU RTH 85 NAM SLEVE 30 25 62 NAM BU RTKE 86 NAM BU RTKE 86 NAM CONF PRE 23 NAM CONF 52 NAM REAL CONF 157 NAM BU 87 NAM BU RRI 89 NAM BU RRV 86 NAM BU RSV 91 NAM BU RTH 85 NAM BU RU 83 NAM BU RV 84 NAM BU RW 84 NAM BU RRC 87 NAM BU RRG 90 NAM BU RRH 91 NAM BU RRI 89 NAM BU RRR 88 NAM BU RRS 90 240 NWET
86. AM CONF PRE LCARTESIAN TRUE LBOUSS FALSE CEQNSYS DUR CIDEAL RSOU CZS FLAT LPERTURB FALSE NVERB 5 amp NAM_CONFn LUSERV TRUE NSV_USER 0 amp NAM DIMn NIMAX 30 NJMAX 30 amp NAM VER GRID NKMAX 16 YZGRID TYPE MANUAL ZDZGRD 70 ZDZTOP 70 ZZMAX STRGRD 1000 ZSTRGRD 0 ZSTRTOP 0 amp NAM_LUNITn CINIFILE ALP ini amp NAM DYNn PRE CPRESOPT NITR 8 XRELAX 1 0 amp NAM LBCn PRE CLBCX 1 CYCL CLBCX 2 CYCL 1 1 1 2 amp NAM_VPROF_PRE CTYPELOC IJGRID NILOC 10 NJLOC 10 36 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE CFUNU ZZZ CFUNV ZZZ LGEOSBAL FALSE GRn PRE CSURF EXTE CH MNHCn PRE LCH INIT FIELD F amp NAM BLANK 3 3 Free format part Each section of the free format part must be introduced by its corresponding key word writen on a separated line There is always a moist variable written in PRE_IDEA1 nam file even in idealized dry cases for which the moist variable should be equal to zero in the PRE IDEAT nam file The produced initial file will always contain a moist variable in CSTN and RSOU cases 3 3 1 Optional Vertical grid keyword ZHAT If the vertical grid generation selector CZGRID TYPE is equal to MANUAL then you must enter just under the namelist part the heights of the vertical ve
87. AM PERT PRE 29 NAM PERT PRE 29 NAM DYNn PRE 26 241 in namelist NAM_DYNn 63 XRELAX HEIGHT FRC in namelist NAM_FRC 56 XRELAX_TIME_FRC in namelist NAM_FRC 56 XRIMKMAX in namelist NAM_DYNn 64 XRPK in namelist NAM_GRID_PRE 26 XSEGLEN in namelist NAM_DYN 54 5 75 in namelist NAM SLEVE 30 XT4DIFF in namelist NAM_DYNn 64 XTADJD in namelist NAM_PARAM_CONVECTn 68 XTADJS in namelist NAM_PARAM_CONVECTn 68 XTNUDGING in namelist NAM_NUDGINGn 66 XTSTEP in namelist NAM_DYNn 63 XVTMAXSURF in namelist NAM_HURR_CONF 161 XWAY in namelist NAM_NESTING 57 XXHATLOC in namelist NAM_VPROFn_PRE 33 XYHATLOC in namelist NAM_VPROFn_PRE 34 Y YDADINIFILE in namelist NAM_LUNIT2_SPA 166 YDADSPAFILE in namelist NAM_LUNIT2_SPA 167 YDOMAIN in namelist NAM_LUNIT2_SPA 166 YNEST in namelist NAM_NEST_PGD 144 YPGDI in namelist NAM_PGD1 144 YPGD2 in namelist NAM_PGD2 144 YPGD3 in namelist NAM_PGD3 144 YPGD4 242 in namelist NAM PGDA 144 YPGD5 in namelist NAM_PGD5 144 YPGD6 in namelist NAM_PGD6 144 YPGD7 in namelist NAM_PGD7 144 YPGD8 in namelist NAM_PGD8 144 YSONFILE in namelist NAM_LUNIT2_SPA 167 YSPAFILE in namelist NAM_LUNIT2_SPA 166 YSPANBR in namelist NAM_LUNIT2_SPA 166 YZGRID_TYPE in namelist NAM_VER_GRID 31 YZOOMFILE in namelist NAM PGDFILE 146 YZOOMNBR in namelist NAM_PGDFILE 146 Z ZDZGRD in namelist NAM_VER_GRID 31 ZDZTOP in namelist
88. AT XZHAT XZHAT 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 TDTCURATDATE TDTCURATIME 444 672 904 1141 1383 1629 1881 2138 2401 2670 2946 3228 3518 3815 4120 4434 4757 5091 5435 5790 6158 6540 6937 7350 7780 8231 8703 9199 9723 10278 10869 11499 12175 12905 13697 14563 15512 16557 17706 18954 20282 21646 22515 1993 18000 TDTEXPATDATE 1993 TDTEXPATIME 79200 TDTMODATDATE 1993 TDTMODATIME 79200 TDTSEGATDATE 1993 TDTSEGATIME 0 79200 YCOMMAND transfer x NIL 46000 53000 91000 80000 40000 90000 70000 90000 90000 90000 40000 60000 10000 20000 50000 50000 80000 00000 10000 80000 90000 70000 30000 10000 80000 20000 30000 80000 80000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 gt 2 gt 2 gt 2 gt 2 23 22 22 22 1 007 CHAPTER 8 THE MESONH FILES 8 2 THE MESO NH FILES amp NAM_LUNITN CCPLFILE 1 CCPLFILE 2 CCPLFILE 3 CCPLFILE 4 CCPLFILE 5 CCPLFILE 6 CCPLFILE 7 CCPLFILE 8 amp NAM CONFN LUSERV T LUSERC T LUSERR T LUSERI T LUSER LUSERH F LUSECI T NSV 0 amp
89. AT logical CEQNSYS 3 characters LFORCING logical NMODEL integer NVERB integer NHALO integer CSPLIT 10 characters LLG logical LINIT_LG logical CINIT_LG 5 characters LNOMIXLG logical CEXP 5 characters CSEG 5 characters e CCONF configuration of all models START for start configuration RESTA for restart configuration YSPLITTING FALSE TRUE PFMOUT FALSE EXPOL e CEQNSYS Equation system resolved by the MESONH model CHAPTER 4 PERFORM A MESONH SIMULATION LHE Lipps and HEmler anelastic system DUR approximated form of the DURran version of the anelastic sytem MAE classical Modified Anelastic Equations but with not any approximation in the momentum equation e LFLAT Switch for zero ororography TRUE no orography zs 0 FALSE the orography is not zero everywhere e LFORCING Switch to use forcing sources TRUE add forcing sources FALSE no forcing sources e NMODEL Number of nested models e NVERB Level of informations on output listing 0 for minimum of prints 5 for intermediate level of prints 10 for maximum of prints e NHALO Size of the halo for parallel distribution This variable is related to computer performance but has no impact on simulation results e CSPLIT Kind of domain splitting for parallel distribution This variable is related to computer performance but has no impact on simulation r
90. BS 0 LOCAL NEXTJOBS 0 or filenames to execute workstation jobs at the end of this job list of jobs executed on local workstation at the end of this job Second argument list of sh command arguments which will run on your workstation with at procedure the pathnames are relative to the current SIMUL directory or absolute job LOCAL NEXTJOBS dirlocal tracei the workstation job will execute cd dirlocal sh tracel dk dk dt dt db H LOCAL NEXTJOBS 0 name of your personal makefile if you don t have one general makefile will be executed on the remote machine according its 05 MKFNAME make_mnh user binary library 0 if none will be searched from your HOME BIBUSER 0 reference binary library BIBMASTER DEFBIBMASTER reference bugfix binary library BIBBUGFIX DEFBIBBUGFIX run with debug options or not run debug 46 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE DEBUG run loading options added to the default ones M f 1 perf LOAD OPT THEHHHHHHHHHHHHHE NEW PROGRAM ZOOM_PGD since masdev4 7 are also available IDEAL CASE MODEL PREP PGD DIAG SPAWNING PREP REAL CASE PREP NEST PGD other names are possible in this case do not forget to fill in NAMELISTFILE MAINPROG O special filename for input namelists for 700 PGD default namelist is PRE ZOOM NAMELISTFILE default special list of variables getting fr
91. BU RTKE in namelist NAM_BU_RTKE 86 LBU RU in namelist NAM BU RU 83 LBU RV in namelist NAM BU RV 84 LBU RW in namelist NAM BU RW 84 LCARTESIAN in namelist NAM CONF PRE 23 LCH CONV LINOX in namelist NAM CH MNHCn 74 LCH CONV SCAV in namelist NAM CH MNHCn 74 LCH EXPLICIT SCAV in namelist NAM CH MNHCn 74 LCH INIT FIELD in namelist NAM CH MNHCn PRE 22 in namelist NAM_CH_MNHCn 74 LCH SURFACE FLUX in namelist NAM CH MNHCn 74 LCH_TUV_ONLINE in namelist NAM CH MNHCn 75 LCHTRANS in namelist NAM PARAM CONVECTn 68 LCLEAR SKY in namelist NAM_PARAM_RADn 70 LCORIO in namelist NAM_DYN 54 LDEEP in namelist NAM PARAM CONVECTn 68 231 LDIAGCONV in namelist NAM PARAM CONVECTn 68 LDOWN in namelist PARAM CONVECTn 68 LDUST in namelist NAM AERO CONF 162 in namelist NAM_DUST_PRE 25 in namelist NAM DUST 53 LFILTERING in namelist NAM HURR CONF 160 LFLAT in namelist _ 52 LFORCING in namelist NAM CONF 24 in namelist NAM CONF 52 LGEOSBAL in namelist NAM VPROFn PRE 32 LGEOST_TH_FRC in namelist NAM_FRC 55 LGEOST_UV_FRC in namelist NAM_FRC 55 LHORELAX_RC in namelist NAM_DYNn 63 LHORELAX_RG in namelist NAM_DYNn 63 LHORELAX_RH in namelist NAM_DYNn 63 LHORELAX_RI in namelist NAM_DYNn 63 LHORELAX_RR in namelist NAM_DYNn 63 LHORELAX_RS in namelist NAM_DYNn 63 LHORELAX_RV in namelist NAM_DYNn 63 LHORELAX_SV in namelist NAM_DYNn 63 LHORELAX_SVAER in namelist
92. D IVAL lt INIL THEN ZVAL FLOAT CIVAL 150 255 190 ZVALMIN MIN ZVAL ZVALMIN ZVALMAX MAX ZVAL ZVALMAX WRITE IUOUT ZLAT ZLON O ZVAL END IF END DO END DO ELSE DO JJ 1 IYSIZE DO JI 1 IXSIZE 195 196 CHAPTER 6 COMPUTE DIAGNOSTICS AFTER A MESO NH SIMULATION 11 11 1 READ IUNIT REC II 2 ZLON ZLONMIN JI ZLONMIN ZLONMAX FLOAT IXSIZE ZLAT ZLATMAX JJ ZLATMAX ZLATMIN FLOAT IYSIZE IVAL ICHAR YTMP2 1 1 2 8 ICHAR YTMP2 2 2 IF IVAL gt O AND IVAL lt INIL THEN ZVAL FLOAT IVAL 0 01 ZVALMIN MIN ZVAL ZVALMIN ZVALMAX MAX ZVAL ZVALMAX WRITE IUOUT ZLAT ZLON O ZVAL END IF END DO END DO ENDIF PRINT VAL min max ZVALMIN ZVALMAX CLOSE IUNIT CLOSE IUOUT END PROGRAM readtarcyl 6 4 References Bolton D 1980 The Computation of Equivalent Potential Temperature Monthly Weather Review 108 1046 1053 Chaboureau J P J P Cammas P Mascart J P Pinty C Claud R Roca and J J Mor crette 2000 Evaluation of a cloud life cycle simulated by Meso NH during FASTEX using METEOSAT radiances and TOVS 31 cloud retrievals Quart J Roy Meteor Soc 126 1735 1750 Hill R J 1980 Refractive index and absorption fluctuations in the infra red caused by temper ature and pressure fluctuations J Opt Soc Am 70 n ircl0 Kunkel B A 1984 Parameterization of droplet terminal velocity and extinction coefficient in fog models J App Meteor 23 34 41 Morcrette
93. D GROUND PARAM FALSE Then you must both define the physiographic and prognostic fields and you must fill the following namelists NAM PGD SCHEMES NAM COVER NAMLISBA if you chose to use the ISBA scheme NAM CH EMIS PGD NAM DUMMY PGD NAM PREP SURF ATM NAM PREP SEAFLUX if you chose to use the SEAFLX scheme NAM PREP WATFLUX if you chose to use the WATFLX scheme NAM PREP TEB if you chose to use the TEB urban scheme NAM PREP ISBA if you chose to use the ISBA scheme 2 You do want to use all the informations contained in a PGD file Then only the prognostic variables must be defined and the following namelists must be filled NAM PREP SURF ATM NAM PREP SEAFLUX if you chose to use the SEAFLX scheme NAM PREP WATFLUX if you chose to use the WATFLX scheme NAM PREP TEB if you chose to use the TEB urban scheme 3 2 THE INPUT THE PRE_IDEA1 NAM FILE 35 e NAM PREP ISBA if you chose to use the ISBA scheme Note that orography either comes from e the input PGD file if any and if LREAD ZS TRUE In this case the atmospheric orography is also set equal to the one in this input PGD file e or from the orography you have defined from the MESONH namelists in this case the surface orography is forced to be equal to the atmosphere orography 3 2 19 Example of namelist part of PREIDEAI1 nam amp NAM REAL PGD CPGD FILE ALPES CPGD LREAD ZS TRUE LREAD GROUND PARAM TRUE amp N
94. D NESTING PREP NEST PGD 149 amp NAM PGD1 YPGD1 PGDFILE 1 amp NAM PGD2 YPGD2 PGDFILE_2 IDAD 1 amp NAM PGD3 YPGD3 PGDFILE_3 IDAD 1 amp NAM PGD4 YPGD4 PGDFILE_4 IDAD 3 amp NAM PGD5 YPGD5 PGDFILE_5 IDAD 2 amp NAM_PGD6 amp NAM PGD7 amp NAM PGD8 amp NAM NEST PGD YNEST 150 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES 5 4 Zoom of a PGD file ZOOM PGD The previous condition on the orography needed when using the gridnesting technique implies that all the PGD files have to be created with PREP PGD and PREP_NEST_PGD programs before beginning the run However the user is not always sure where and when initialize the inner models To avoid to set exactly the domain of inner models at the PREP PGD step one solution is to make PGD file on larger domain and then zoom it on the part of the domain of interest when knowned with the following program ZOOM PGD Then the output PGD file is used as PGD file for the interpolations of atmospheric fields with SPAWNING and PREP REAL CASE programs ZOOM PGD is run with the MESO NH procedure prepmodel In the file prepmodelrc the input host directories and login control variables refer to the input PGD file The other control variables to initialize specifically in this file are e MAINPROG ZOOM PGD e NAMELISTFILE PRE_ZOOM e LISTGET CPGDFILE The namelist file PRE ZOOMI nam contains 2 namelist
95. DMR ZUVTHLMR Except for the STANDARD kind the first letter of KIND represents the kind of altitude variable P for pressure and Z for height the second and third letters represent the kind of wind variables U for zonal wind V for meridian wind the fourth fifth and sixth letters represent the kind of temperature variable THV for virtual potential temperature THD for dry potential temperature and THL for liquid potential temperature the seventh and heighth letters represent the kind of moist variable HU for relative humidity and MR for vapor mixing ratio In case of STANDARD kind the altitude variable is the pressure the wind variables are direction and force of wind the temperature variable is the temperature and the moist variable is the due point temperature HEIGHT of GROUND LEVEL real in meters e PRESSURE at GROUND LEVEL real in Pascal e a TEMPERATURE variable at GROUND LEVEL real in Kelvin a MOIST variable at GROUND LEVEL e NUMBER WIND data LEVELS integer e level 1 ALTITUDE variable first WIND variable second WIND variable at wind level 1 the lowest wind level e level 2 ALTITUDE variable first WIND variable second WIND variable 3 8 FREE FORMAT PART 39 uppermost wind level ALTITUDE variable first WIND variable second WIND variable NUMBER of mass data LEVELS integer Note that this number includes the ground level i e the first level That
96. ELTAX in namelist NAM_GRIDH_PRE 27 XDELTAY in namelist NAM_GRIDH_PRE 27 XDTCONV in namelist NAM_PARAM_CONVECTn 67 XDTRAD in namelist NAM_PARAM_RADn 68 XDTRAD_CLONLY in namelist NAM_PARAM_RADn 69 XFMOUT in namelist NAM_FMOUT 55 XFUDG in namelist NAM_PARAM_RADn 70 XHMAX in namelist NAM_GRIDH_PRE 27 XIMPL in namelist NAM_TURBn 71 XINIRADIUSI in namelist 162 INDEX in namelist XINIRADIUSJ in namelist in namelist XINISIGI in namelist in namelist XINISIGJ in namelist in namelist XLAMBDA in namelist XLATO in namelist XLATBOG in namelist XLATGUESS in namelist XLATLOC in namelist XLATORI in namelist XLENI in namelist XLEN2 in namelist XLONO in namelist XLONBOG in namelist XLONGUESS in namelist XLONLOC in namelist XLONORI in namelist XRADGUESS in namelist XRADWINDSURF in namelist XRADX in namelist XRADY in namelist XRADZ in namelist XRELAX in namelist NAM CH ORILAM 76 NAM AERO CONF 162 NAM CH ORILAM 76 NAM AERO CONF 162 NAM CH ORILAM 77 NAM AERO CONF 162 NAM CH ORILAM 77 NAM HURR CONF 160 _ 26 NAM HURR CONF 161 NAM HURR CONF 160 NAM VPROFn PRE 33 _ 26 NAM VER GRID 32 159 NAM VER GRID 32 159 _ 26 NAM HURR CONF 161 NAM HURR CONF 160 NAM VPROFn PRE 33 _ 26 NAM HURR CONF 160 NAM HURR CONF 161 NAM PERT PRE 29 N
97. ETHRI wet growth of hail ICE4 integer Hy gt 4 U D O 42 THE INPUT EXSEGS N NAM FILE 4 2 39 Namelist NAM BU RRS budget for snow Fortran type default value LBU_RRS budget switch logical NASSERS time filter Asselin integer NNESTRS nesting integer NADVRS total advection integer NADVXRS advection along x integer NADVYRS advection along y integer NADVZRS advection along z integer NFRCRS forcing integer NDIFRS numerical diffusion integer NRELRS relaxation integer NNEGARS negative correction integer NSEDIRS sedimentation ICE3 or ICE4 integer NDEPSRS deposition on snow ICE3 or ICE4 integer NAGGSRS aggregation of snow ICE3 or ICE4 integer NAUTSRS autoconversion of ice ICE3 or ICE4 integer NRIMRS riming of cloudwater ICE3 or ICE4 integer NACCRS accretion of rainwater ICE3 or ICE4 integer NCMELRS conversion MeLTing ICE3 or ICE4 integer NWETGRS wet growth of graupel ICE3 or ICE4 integer NDRYGRS dry growth of graupel ICE3 or ICE4 integer NWETHRS wet growth of hail ICE4 integer Hy gt in U 9 4 2 40 Namelist NAM BU budget for graupel Fortran type default value LBU_RRG NASSERG NNESTRG NADVRG NADVXRG NADVYRG NADVZRG NFRCRG NDIFRG NRELRG NNEGARG NSEDIRG NSFRRG NDEPGRG NRIMRG NCMELRG NCFRZRG NWETGRG NDRYGRG NGMLTRG NWETHRG budget sw
98. FCT2ND NLITER integer 2 It contains the different advection schemes for dynamic variables u v and w scalar me teorological variables temperature water substances and tracers used by the model n They are included in the declarative module MODD_ADVn e CUVW SCHEME Advection scheme used for horizontal and vertical velocities The following options are possible 42 THE INPUT EXSEGS N NAM FILE 63 CEN2ND 2nd order advection scheme CENtred on space and time It does NOT guarantee the sign preservation CENATH 4th order advection scheme CENtred on space and time It does NOT guarantee the sign preservation e SCHEME Advection scheme used for the following METeorological vari ables temperature water substances and TKE The following options are possible see the Scientific Documentation for more details CEN2ND 2nd order advection scheme CENtred on space and time It does NOT guarantee the sign preservation CENATH 4th order advection scheme CENtred on space and time It does NOT guarantee the sign preservation FCT2ND 2nd order advection scheme CENtred on space and time It is POSITIVE definite MPDATA 2nd order advection scheme uncentred on space and time It is POSITIVE definite PPM_00 PPM advection scheme without constraint PPM 01 Monotonic version of PPM It is POSITIVE definite e CSV ADV 5 Ad
99. G e NAMELISTFILE default e LISTGET default 5 8 2 The input SPAWNI nam file The order of namelists is free and unset namelists can be ommited 1 Namelist NAM GRID2 SPA manual definition of domain IXOR integer 1 IYOR integer 1 IXSIZE integer file 1 domain IYSIZE integer file 1 domain IDXRATIO integer 1 IDYRATIO integer 1 GBAL ONLY logical FALSE e IXOR first point I index according to the file 1 grid left to and out of the new physical domain 170 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES IYOR first point J index according to the file 1 grid under and out of the new physical domain IXSIZE number of grid points in I direction according to file 1 grid recovered by the new domain It must only be factor of 2 3 or 5 IYSIZE number of grid points in J direction according to file 1 grid recovered by the new domain It must only be factor of 2 3 or 5 IDXRATIO resolution factor in I direction between the file 1 grid and the new grid It must only be factor of 2 3 or 5 IDYRATIO resolution factor in J direction between the file 1 grid and the new grid It must only be factor of 2 3 or 5 ONLY switch to enforce anelastic constraint only The spawned file have the same caracteristics than the CINIFILE one 2 Namelist NAM LUNIT2 SPA file names CINIFILE character len 28 INIFILE YDOMAIN character 28 none YSPAFILE character 28 none YSPANBR character l
100. G DPR 7 dyn prod by resolved fluctuations lt ul ri 2 6 gt lt gz w m gt BU_THLR SBG TR subgrid turbulent transport residual of budget 27977 gt BU THLR SBC RESI zt must be small 44 BUDGET OF RESOLVED SUBGRID TURBULENT QUANTITIES 121 field notation in processus dim comments ite ame peste d ee ea of lt gt eee ER e by mean flow lt gt BU THLR RES FORC z t advection by large scale W forcing i wn nal or by mean gradients E BU THLR RES TR resolved transport of resolved flux BU THLR RES SBGT 7 1 sink due to subgrid turbulence EN of lt 0j gt BULTHLR RES NUMD numerical diffusion against 2Ax relaxation of 6 gt BU THLR RES RELA 2way nesting of Or gt BU_THLR RES NEST z t average from smaller 8 miscellaneous THLR RES MISC ref pressure term residual of budget of lt gt z t must be zero must be zero Zero lt gt neglected opposite of neglected in turb scheme tendency of lt w r gt lt gt lt br gt neglected advection by mean flow lt Ua a Ort gt neglected advection by resolved flow 122 CHAPTER 4 PERFORM A MESONH SIMULATION 4 4 6 Budget of total water variance terms of the equation of lt gt lt gt
101. GD file TRUE to use the data read in the PGD FILE FALSE to use XUNIF COVER idealized homogeneous values given in the namelist NAM COVER from the externalized surface and scratch the PGD FILE data e LREAD ZS switch to use or not the orography parameters read in the PGD file TRUE to use the data read in the PGD FILE FALSE to use an idealized orography given in the namelist NAM GRIDH PRE and scratch the PGD FILE data 3 2 15 Namelist NAM SLEVE smoothed orography for Sleve coordinate Fortran type default value NSLEVE integer 12 XSMOOTH ZS real XUNDEF e NSLEVE number of iteration for computation of smooth orography XSMOOTHLZS optional uniform smooth orography 3 2 16 Namelist NAM VER GRID contains vertical grid definition There are three ways to compute the vertical grid as in PREP REAL CASE 1 constant grid mesh only the number of levels NKMAX and the grid mesh sizes ZDZGRD and ZDZTOP are used These must be equal The type of grid YZGRID_TYPE is set to FUNCTN 3 2 THE INPUT THE PRE_IDEA1 NAM FILE 31 2 two layers are defined with constant stretching in each of these the grid mesh sizes being given near the ground and at top of the model It is possible that the top grid size is never reached if the number of points is not enough for the prescribed stretchings type of grid YZGRID is also set to FUNCTN 3 the levels are given by the user The
102. GTH in namelist NAM_BU_RTH 85 NWETHRC in namelist NAM_BU_RRC 87 NWETHRG in namelist NAM_BU_RRG 90 NWETHRH in namelist NAM BU RRH 91 NWETHRI in namelist NAM BU RRI 89 NWETHRR in namelist NAM BU RRR 88 NWETHRS in namelist NAM BU RRS 90 NWETHTH in namelist NAM_BU_RTH 85 NXOR in namelist NAM MESONH DOM 147 NYOR in namelist NAM MESONH DOM 147 X XALKTOP in namelist NAM DYN 54 XALZBOT in namelist NAM_DYN 54 XAMPLIRV in namelist NAM_PERT_PRE 29 XAMPLITH in namelist NAM_PERT_PRE 29 XAMPLIUV in namelist NAM_PERT_PRE 29 XAMPLIWH in namelist NAM_PERT_PRE 29 XASSELIN in namelist NAM_DYN 54 XASSELIN SV in namelist NAM_DYN 54 XAX in namelist NAM_GRIDH_PRE 27 XAY in namelist NAM_GRIDH_PRE 27 XBETA in namelist NAM GRID PRE 26 XBOXWIND in namelist NAM HURR CONF 160 XBULEN in namelist NAM_BUDGET 81 INDEX XBUWRI in namelist NAM BUDGET 81 XCELMAX in namelist NAM TURB CLOUD 60 XCEI_MIN in namelist NAM_TURB_CLOUD 60 XCENTERZ in namelist NAM_PERT_PRE 29 XCH_TS1D_TSTEP in namelist NAM CH MNHCn 75 XCH_TUV_ALBNEW in namelist NAM CH MNHCn 75 TUV DOBNEW in namelist NAM CH MNHCn 75 XCH_TUV_TUPDATE in namelist NAM CH MNHCn 75 XCOEF_AMPL_SAT in namelist NAM_TURB_CLOUD 60 XCOEFRADIMAX in namelist NAM_CH_ORILAM 77 XCOEFRADIMIN in namelist NAM_CH_ORILAM 77 XCOEFRADJMAX in namelist NAM_CH_ORILAM 77 XCOEFRADJMIN in namelist NAM_CH_ORILAM 77 XCPHASE in namelist NAM_LBCn 64 XD
103. ICS AFTER A MESO NH SIMULATION OPEN IUNIT FILE YFNAM IOSTAT IO IF IO 0 THEN PRINT PROBLEM WHILE OPENING STOP ELSE PRINT OPEN TRIM YFNAM ENDIF YLBLCH UNDX XXX DO READ IUNIT A END 99 YLINE SELECT CASE YLINE 1 6 CASE SATIMG YSAT YLINE 8 20 IF YSAT 1 8 meteosat YLBLCH 1 4 MET YSAT 10 10 IF YSAT 1 10 meteosat08 YLBLCH 1 4 MSG1 IF YSAT 1 4 goes YLBLCH 1 4 GO YSAT 5 6 CASEC NBYTE BACKSPACE IUNIT READ IUNIT 26 YLINE 1 6 NBYTE CASE XSIZE BACKSPACE IUNIT READ IUNIT 26 YLINE 1 6 IXSIZE CASE YSIZE BACKSPACE IUNIT READ IUNIT 26 YLINE 1 6 IYSIZE CASE C LATMIN YMEM TRIM YMEM TRIM YLINE BACKSPACE IUNIT READ IUNIT 27 YLINE 1 7 ZLATMIN CASE C LATMAX YMEM TRIM YMEM TRIM YLINE BACKSPACE IUNIT READ IUNIT 27 YLINE 1 7 ZLATMAX CASE C LONMIN YMEM TRIM YMEM TRIM YLINE BACKSPACE IUNIT READ IUNIT 27 YLINE 1 7 ZLONMIN CASE CO LONMAX YMEM TRIM YMEM TRIM YLINE BACKSPACE IUNIT READ IUNIT 27 YLINE 1 7 ZLONMAX END SELECT IF YLINE 1 3 ID THEN IF YLBLCH 1 3 MET THEN IF YLINE 4 5 ai YLBLCH 6 8 IR IF YLINE 4 5 aw YLBLCH 6 8 WV ELSEIF YLBLCH 1 3 MSG THEN YLBLCH 6 8 7 9 ELSEIF YLBLCH 1 2 G0 THEN IF YLINE 4 4 227 YLBLCH 6 8 039 IF YLINE 4 4
104. ITE ILUOUT CFL NUMBER ZCFL WRITE ILUOUT MAXIMUM HORIZONTAL WIND ZWIND MAX END IF END SUBROUTINE CFL FILE model n f90 HHHHHHSpL SUBROUTINE MODEL_n KTCOUNT OEXIT HEHEHE HEHEHE HEHEHE HE HH HE RETE HE SERE SERERE SERERE SERERE RE ERETERERERERE MODEL_n monitor of the model version n USE MODD BLANK USE MODI CFL IMPLICIT NONE Ix 2 25bis CFL control 7 4 AN ILLUSTRATIVE EXAMPLE OF A SOURCE COMPILATION 207 IF NOT LDUMMY1 THEN CALL CFL XUT XVT XDXHAT XDYHAT XTSTEP CLUOUT TDTCUR END IF END SUBROUTINE MODEL n The first step is to create a directory which will contain the RCS libraries and to go in this directory mkdir SOURCE cd SOURCE Create the new source files cfl f90 in our case in this directory with any editor Modify the already existing source files first take it from the Master or Bugfix source libraries for our case for MESONH sources adiab or MESONH bugfix mastern nh takeout modeln f90 MASDEV4_2 mesonh sources adiab the mesonh sources adiab is the answer to the question asked by the nh takeout macro to determine the DIRSRC directory Then create your RCS library nh create modeln f90 EXAMPLE nh create cfl f90 EXAMPLE the is the answer to the question asked by the nh create macro to determine the DIRSRC directory Then the Meso NH environment variables must be checked refer to the previous exam
105. J J 1991 Radiation and cloud radiative properties in the European Center for Medium Weather Forecasts forecasting system J Geophys Res 96 9121 9132 Saunders R M Matricardi P Brunel S English P Bauer U O Keeffe P Francis and Rayer 2005 RT TOV 8 science and validation report NWP SAF Report 41 pages Tech rep Chapter 7 Modify the Fortran sources 7 1 Presentation of the different libraries The source code of the model is managed by the UNIX standard source manager RCS Revision Control System present on Linux computers With this source manager you are able to get the source either from a reference library 1 the Master library or the Bugfix library or from a personal library Master and Bugfix sources build a safe set of Fortran 90 mains and subroutines modified and updated by the administrator of the Meso NH sources We have seen in the previous chapters how to use the Meso NH model as it stands by collecting these binary objects previously compiled on the remote_host For your own purposes you can add new sources or modify the Master sources by building your own library the User s library Eventually some of these modifications may be incorpo rated in the next release of the Master library after a merging with modifications from other users In this case you will have to manipulate new objects with the procedures the User s library on the local_host and its binary counte
106. JMIN factor to compute minimum value of mean radius mode J accumula tion mode RM XCOEFRADJMIN XINIRADIUSJ e CMINERAL type of parameterization for mineral gas particle balance Possible values are CMINERAL ARES ARES parameterization non vectorized CMINERAL NARES neuronal network of ARES vectorized CMINERAL ISPIA ISORROPIA parameterization non vectorized CMINERAL TABUL tabulation of ISORROPIA vectorized CMINERAL EQSAM EQSAM parameterization vectorized e CORGANIC type of parameterization for organic gas particle balance To activate or ganic parameterization it is necessary to use a chemical scheme capable to form secondary organic aerosol i e RELACS2 or CACM Possible values are CORGANIC PUN PUN parameterization CORGANIC MPMPO MPMPO non vectorized e CNUCLEATION type of parameterization for nucleation formation of new particle from sulphates Possible values are CNUCLEATION KULMALA KULMALA parameterization CNUCLEATION KERMINEN KERMINEN parameterization e Convective scavenging is activated with LCH CONV SCAV in NAM CH MNHCn 42 THE INPUT EXSEGS N NAM FILE 4 2 24 8l Namelist NAM CH SOLVERn control stiff solvers for modeln Fortran type default value CSOLVER NSSA NSSAINDEX XRTOL XATOL NRELAB NPED NMAXORD LPETZLD CMETHOD CNORM NTRACE XALPHA XSLOW XFAST NQSSAITER XDTMIN XDTMAX XDTFIRST 32 char
107. Ju cem by mean flow W orc lt i0 gt BU WTHL RES FORC 24 advection by large scale W forcing lt 4 gt 2 lt 0 gt BU_WTHL RES DP z t dyn prod by mean gradients lt gt 2 lt w gt mem flux by itself zm LIAE BUWTAL RES PRES zi pressureconlations BU WTAL RES TP BU WTHL RES SBGT z t sink due to subgrid turbulence T numerical diffusion of lt 400 gt BU WTHL RES NUMD 2 1 numerical diffusion against 2 relaxation of lt gt BUTWTHL RES RELA zt sponge layer relaxation E nesting of lt i0 gt U_WTHL RES NEST 2 average from smaller nested models miscellaneous WTHL RES MISC z t ref pressure term curvature term microphysics radiation residual of budget of lt gt BU WTHL RES RESI zt mustbezero wi BU WTHL NSG TEND 2 6 neglected opposite of neglected in turb scheme tendency lt gt 2 lt gt neglected advection by mean flow gt neglected advection by resolved flow BU WTHL NSG DPGW neglected dyn prod terms terms due to hor 0 gradients BULWTHL NSG 2 1 other neglected dyn prod terms 116 CHAPTER 4 PERFORM A MESONH SIMULATION 4 4 3 Budget of total liquid temperature variance A l terms of the equation of lt gt lt gt are computed and stored in the diachronic group BU TH
108. L CRAD ECMWF CGROUND ISBA CDCONV KAFR PARAM RADn XDTRAD 300 XDTRAD_CLONLY 150 NRAD_COLNBR 400 CAER TANR CLW MORC amp NAM_PARAM_CONVECTn XDTCONV 300 NICE 1 LREFRESH_ALL T LDEEP T LSHAL T LDOWN T amp NAM_PARAM_GROUNDN CROUGH ZO1D amp NAM LBCn 2 0PEN CLBCY 2x OPEN TURBn CTURBLEN BL89 CTURBDIM 1DIM LSUBG COND T LSUBG_AUCV T LTURB_DIAG FALSE LTURB_FLX FALSE LSIG CONV T LSIGMAS F 45 AN ILLUSTRATIVE EXAMPLE OF A MESONH SIMULATION 133 CONF RESTA NVERB 2 NMODEL 2 CEXP 16736 CSEG 12B18 DYN XSEGLEN 800 LCORIO T LNUMDIFF LSTEADYLS F XALKTOP 0 001 XALZBOT amp NAM NESTING NDAD 2 1 NDTRATIO 2 3 XWAY 2 2 5 amp NAM FMOUT XFMOUT 1 1 10800 XFMOUT 1 2 21600 XFMOUT 2 1 10800 XFMOUT 2 2 21600 14500 amp NAM ISBAn CRUNOFF WSAT CSCOND NP89 CALBEDO DRY CCiDRY CSOILFRZ DEF CDIFSFCOND DEF CSNOWRES DEF CROUGH Z04D amp NAM_SEAFLUXn CSEA_ALB UNIF CSEA_FLUX DIRECT amp NAM_DIAG_SURFn amp NAM_DIAG_ISBAn amp NAM_DIAG_SURF_ATMn The five latest namelists are for the externalised surface e To execute the preparative job on the local_host enter prepmodel Then answer to the questions asked by the procedure for the simulation directory if the environme
109. L2 All comments made for the total Tke equation are valid here ADVM ADVR DPM 0 lt 0 gt 9 lt 02 gt CURSUM Laer lt tap 0r gt 2 lt ul gi gt A 20 06 5 2 u 0 gt lt u 0 gt e Z SE LL DISS DPR TR ADV DP 2 2 lt h gt 2 6 gt lt Ua gt lt 0 gt 2 lt gt d zug 22 6 20 0 gt EH ard TR SBGT field notation in processus dim comments 2 lt 0q gt pas BUNTHL SBGDPM zt d pod mem gadi gt FBUOWTHL SBG DP R zt prod by resolved fuctuations lt v PBU WTHE SBGTR turbulent transport lt gt o o BUWTHL SEGDISS t dissipation residual of budget of lt WA BU WTHL SBG RESI zt mwthesmdl 2 lt 6 gt BU_WTHL RES TEND 7 opposite of tendency of lt 6 gt WHHL RES at advection by mem fow BU ic a i MN large scale W forcing by mean gradients resolved variance 44 BUDGET OF RESOLVED SUBGRID TURBULENT QUANTITIES 117 field notation in processus dim comments diac file name lt 20 d ul 9 gt U numerical diffusion of lt 6 gt BU U J WTHL RES SBGT WTHL RES NUMD 2way nesting of lt 6 gt BU WTHL RES NEST miscellaneous BU WTHL
110. M TRUE e LVARSIGI switch to activate variable standard deviation for mode I Aitken e LVARSIGJ switch to activate variable standard deviation for mode J accumulation e LSEDIMAERO switch to activate aerosol sedimentation e LTHERMIJ switch to compute gas particles thermodynamic balance with separate modes if TRUE or for all aerosol mass if FALSE e XINIRADIUSI to switch on the initialization of mean radius mode I Aitken mode of the distribution in micrometers e XINIRADIUSJ to switch on the initialization of mean radius mode J accumulation mode of the distribution in micrometers e CRGUNIT type of mean radius given in namelist Default is for a mass spectral dis tribution XINIRADIUSI and XINIRADIUSJ has been converted into a mean radius in number IF CRGUNIT MASS then the mean radius need to be given for a number spectral distribution no conversion e XINISIGI value of standard deviation for mode I Aitken mode 80 CHAPTER 4 PERFORM A MESONH SIMULATION XINISIGJ value of standard deviation for mode J accumulation mode e XCOEFRADIMAX factor to compute maximum value of mean radius mode I Aitken mode 1 XCOEFRADIMAX XINIRADIUSI e XCOEFRADJMAX factor to compute maximum value of mean radius mode J accumu lation mode XCOEFRADJMAX XINIRADIUSJ e XCOEFRADIMIN factor to compute minimum value of mean radius mode I Aitken mode RP XCOEFRADIMIN XINIRADIUSI e XCOEFRAD
111. M BU RU 83 NAM BU RV 84 NAM BU RW 84 NAM BU RU 83 NAM BU RV 84 NAM BU RW 84 NAM NESTING 57 NAM BU RRC 87 NAM BU RRI 89 NAM BU RRV 86 NAM BU RSV 91 NAM BU RTH 85 NAM BU RRG 90 in namelist NDEPGTH in namelist NDEPIRC in namelist NDEPSRS in namelist NDEPSRV in namelist NDEPSTH in namelist NDIAG FILT in namelist NDIFRC in namelist NDIFRG in namelist NDIFRH in namelist NDIFRI in namelist NDIFRR in namelist NDIFRS in namelist NDIFRV in namelist NDIFSV in namelist NDIFTH in namelist NDIFTKE in namelist NDIFU in namelist NDIFV in namelist NDIFW in namelist NDISSHTH in namelist NDISSTKE in namelist NDPTKE in namelist NDRYGRC in namelist NDRYGRG NAM BU RRV 86 NAM BU RTH 85 NAM BU RRC 87 NAM BU RRS 90 NAM BU RRV 86 NAM BU RTH 85 NAM BU RRC 87 NAM BU RRG 90 NAM BU RRH 91 NAM BU RRI 89 NAM BU RRR 88 NAM BU RRS 90 NAM BU RRV 86 NAM BU RSV 91 NAM BU RTH 85 NAM BU RTKE 86 NAM BU RU 83 NAM BU RV 84 NAM BU RW 84 NAM BU RTH 85 NAM BU RTKE 86 NAM BU RTKE 86 NAM BU RRC 87 INDEX NAM HURR CONF 161 INDEX in namelist NAM_BU_RRG 90 NDRYGRI NDRYGRR NDRYGRS NDRYGTH NDTRATIO NENSM NEXPX in namelist NFRCRC in namelist NFRCRG in namelist NFRCRH in namelist NFRCRI in namelist NFRCRR in namelist NFRCRS in namelist NFRCRV in namelist NFRCSV in na
112. MASTER DEFBIBMASTER reference bugfix binary library BIBBUGFIX DEFBIBBUGFIX run with debug options or not run debug DEBUG run loading options added to the default ones M f 1 perf LOAD OPT HHHHHHHHHHHHHHH NEW PROGRAM 200 PGD since masdev4 7 are also available IDEAL CASE MODEL PREP PGD DIAG SPAWNING PREP REAL CAS PREP_NEST_PGD other names are possible in this case do not forget to fill in NAMELISTFILE MAINPROG 0 special filename for input namelists for ZOOM_PGD default namelist is PRE_ZOOM NAMELISTFILE default special list of variables getting from NAMELISTFILE 130 CHAPTER 4 PERFORM A MESONH SIMULATION LISTGET default 4 input files location of the input FM files on execution machine or storage machine HOME INDIR supc archiv INHOST supc to get the input FM files directory name starting at HOME if it begins with home at workdir if it begins with work indicate one or several directories between double quotes INDIR work INOUTFILES user name on which the get will be executed O LOGNAME INLOGIN 0 Overview of archiv case INHOST archiv INDIR mxxx mxxx007 CHEMIN mgrp999 NEXTDOOR INLOGIN 0 OR INHOST archiv INDIR NEXTDOOR INLOGIN mrgp999 output files storage of the output FM files on execution machine or storage machine supc archiv RMACH aerosv2
113. ME 3 SST_2001062300 CDUMMY AREA 3 gt SEA CDUMMY_ATYPE 3 ART CDUMMY_FILE 3 5Tn2001062300 dat CDUMMY_FILETYPE 3 ASCLLV CDUMMY_NAME 4 SST_2001062400 CDUMMY_AREA 4 gt SEA CDUMMY_ATYPE 4 ART CDUMMY_FILE 4 5Tn2001062400 dat CDUMMY_FILETYPE 4 ASCLLV CDUMMY_NAME 5 SST_2001062500 CDUMMY_AREA 5 SEA CDUMMY_ATYPE 5 ARI CDUMMY_FILE 5 55Tn2001062500 dat CDUMMY_FILETYPE 5 ASCLLV CDUMMY_NAME 6 SST_2001062600 5 2 CREATION OF MESO NH PHYSIOGRAPHIC DATA FILE PREP_PGD 147 CDUMMY AREA 6 SEA CDUMMY_ATYPE 6 ARI CDUMMY_FILE 6 5Tn2001062600 dat CDUMMY_FILETYPE 6 ASCLLV Another PREP_PGD run for the father PGD file with all the namelist variables amp NAM PGDFILE CPGDFILE PGD_AMMA1_10km_m46_b1 amp NAM PGD SCHEMES CNATURE ISBA CSEA SEAFLX CWATER WATFLX CTOWN TEB amp NAM PGD GRID CGRID CONF PRO0J amp NAM CONF PROJ 13 XLONO 1 5 XRPK 0 2249510543 0 CONF PROJ XLATCEN 13 XLONCEN 1 5 NIMAX 324 NJMAX 240 XDX 10000 XDY 10000 amp NAM COVER YCOVER ecoclimats_v2 YFILETYPE DIRECT LRM_TOWN FALSE amp NAM_ZS YZS gtopo30 YFILETYPE DIRECT COROGTYPE AVG XENV 0 NZSFILTER 1 amp NAM_ISBA NPATCH 12 CISBA 3 L CPHOTO AGS NGROUND_LAYER 3 YCLAY clay_fao YCLA
114. MYOWNL version 1 1 of source f90 associated to MYOWNL nh extract source f90 MYOWNL version 1 1 extracted for modification next version will be 1 2 After this extraction you have a file source f90 which contains the source corresponding to version 1 1 Therefore you can edit this file and modify it as you want nh register source f90 MYOWNL version 1 2 created and associated to MYOWNL Note that the file source f90 which appeared in your directory when the nh extract was performed has now disappeared From now on if the user needs to look at the version 1 1 without requiring any modifica tions he must do nh takeout source f90 MYOWNL 1 1 But let s let the user go on making a new version 1 3 nh extract source f90 MYOWNL Therefore he extracts the last available version i e version 1 2 and by modifying source f90 with the editor he makes the new version 1 3 registered by the following nh register operation nh register source f90 MYOWNL However the user now recognizes an error and needs to remove completly version 1 3 in this case he must do nh unregister source f90 MYOWNL version 1 2 is now again the last default version for MYOWNL The user can also create a new branch starting from 1 1 that will be parallel to the main trunk 221 222 APPENDIX A AN ILLUSTRATIVE EXAMPLE OF USE OF PROCEDURES nh extract source f90 MYOWNL 1 1 where the new version to be generated will be 1 1 1 1 If after careful thoughts the user
115. N If HPARAM_CCN CPB then the initial CCN characteristics are given the or AER format In the CCN case the parameters XCHEN XKHEN XMUHEN and XBETAHEN must be given while it is the case for XCONC CCN XR _ XLOGSIG CCN XFSOLUB CCN and XACTEMP CON if the AER option is chosen CCN The aerosols are directly characterized by their activation spectrum Noow s in the form C s or C s F y 5 E 1 6882 AER aerosols are particles which are characterized by a lognormal distribution law in the form In r 7 2In o with distribution parameters 7 is the geometric mean radius the geometric standard deviation and N the total particle number by their solubility and by their activation temperature as described by Cohard et al JGR 2000 e HTYPE Aerosol type or C if HPARAM_CCN CPB and HINL CCN AER is chosen M NaCl composition large size maritime aerosols C NH4 2S0 composition small size continental aerosols e XCHEN Parameter C the generic activation spectrum Noow s e XKHEN Parameter k in the generic activation spectrum 8 XMUHEN Parameter the hypergeometric function of the CPB form of the activation spectrum 8 42 THE INPUT EXSEGS N NAM FILE 61 e XBETAHEN Parameter 8 in the hypergeometric function of the CPB form of the a
116. N user on the INHOST machine of the input mesonh files 02 the LOGNAME on the remote host e OUTHOST output mesonh files will be on execution remote host storage file machine or workstation supc archiv workstation name e OUTDIR directory of the output mesonh files as INDIR 18 CHAPTER 2 THE MESONH PROCEDURES EXCEPT PREPSOURCE CCPLFILE PREP PGD PRE_PGD1 nam YZS YCLAY outprep pgd PREP NEST PGD PRE NEST PGDI nam YPGDI outprepnest PREP REAL CASE PRE REALI nam HATMFILE outprepreal SPAWNING SPAWNI nam CINIFILE outprepspawn YDOMAIN YDADINIFILE YDADSPAFILE ZOOM PGD PRE ZOOMI nam CPGDFILE MY_PROG MY_NAM1 nam outprepmodel NAMELISTFILE MY_NAM LISTGET Table 2 1 The different MesoNH programs and Namelist files associated e SUBMIT_NEXTJOBS name of one or several jobs to submit to remote host at the end of this job pathnames relative to the current SIMUL tosupcrc must be present in the same directory 0 filename e LOCAL_NEXTJOBS name of jobs to execute on local host at the end of this job with at procedure pathnames relative to the current SIMUL or absolute 0 dirlocal filename The last two control parameters allow to link several jobs together the first one is executed on the remote host output of prepmodel the other ones are either executed on the remote host e g a model simulation after the preparation of the initial file or on the l
117. NAM DYNN XTSTEP 7 500 CPRESOPT RICHA NITR XRELAX 1 0000 LHO RELAX F LVE RELAX NRIMX NRIMY 4 XRIMKMAX 0 00050000 XT4DIFF 1000 0000 amp NAM ADVN CMET ADV SCHEME FCT2ND CSV ADV SCHEME FCT2ND CFV ADV SCHEME MPDATA NLITER 2 4 gt 4 gt gt gt 219 S T LUSERG T PARAMN CTURB TKEL CRAD NONE CDRAG NONE CGROUND FLUX CCLOUD ICE3 CDCONV NONE PARAM RADN XDTRAD 60 XDTRAD_CLONLY 60 LCLEAR_SKY F NSPOT 1 NRAD_COLNBR 1000 amp NAM_PARAM_CONVECTN XDTCONV 300 NICE 1 amp NAM PARAM GROUNDN CROUGH 204 LBCN CLBCX 2 OPEN CLBCY 2 CYCL NLBLX 2 1 NLBLY 2 1 XCPHASE 20 TURBN 0 E 0 CTURBLEN DELT CTURBDIM 3DIM LTURB_FLX F LTURB_DIAG F LSUBG_COND F amp NAM CH MNHCN LUSECHEM F LCH INIT FIELD LCH SURFACE FLUX CCH INIT FIELD OPT NONE CCH SURFACE FLUX OPT NONE gt CCH_TDISCRETIZATION SP CONF START 12 11 F LFLAT T NMODEL NVERB 1 CEXP COPTE CSEG NEWIC LFORCING F amp NAM DYN XSEGLEN 28807 5 XASSELIN 0 2000000000000002 LCORI LNUMDIFF T LSTEADYLS T XALKTOP 5 000000000000004E 3 XALZB amp NAM_NESTING NDAD 1 2 3 4 5 6 7 8 NDTRATIO 8 FMOUT XFMOUT 8 3600 8 7200 8 10800 8 14400 8 BUDGET CBUTYPE
118. NAM_DYNn 63 LHORELAX_SVC1R3 in namelist NAM_DYNn 63 LHORELAX_SVC2R2 in namelist NAM_DYNn 63 LHORELAX_SVCHEM in namelist NAM_DYNn 63 LHORELAX_SVDST in namelist NAM_DYNn 63 LHORELAX_SVELEC 232 in namelist DYNn 63 LHORELAX SVLG in namelist NAM_DYNn 63 LHORELAX TKE in namelist NAM_DYNn 63 LHORELAX UVWTH in namelist NAM_DYNn 63 LINIT LG in namelist _ 52 LITRADJ in namelist NAM_DYNn 63 LLG in namelist _ 52 LMASKLANDSEA in namelist NAM SERIES 79 LNOMIXLG in namelist NAM_CONF 53 LNUDGING in namelist NAM NUDGINGn 66 LNUMDIFF in namelist NAM DYN 54 LORILAM in namelist NAM AERO CONF 162 in namelist NAM CH MNHCn PRE 22 in namelist NAM CH ORILAM 76 LPACK in namelist NAM_CONF_PRE 23 LPERTURB in namelist NAM_CONF_PRE 24 LPGROUND FRC in namelist NAM_FRC 57 LREAD GROUND PARAM in namelist NAM_REAL_PGD 30 LREAD ZS in namelist NAM REAL PGD 30 LREFRESH ALL in namelist NAM_PARAM_CONVECTn 67 LRELAX_THRV_FRC in namelist NAM_FRC 56 LRELAX_UV_FRC in namelist NAM_FRC 56 LRMCO1 in namelist NAM_TURBn 72 LSEDIC in namelist NAM PARAM ICE 59 LSEDIMAERO in namelist NAM CH ORILAM 76 LSEDIMDUST INDEX in namelist NAM DUST 53 LSERIES in namelist NAM SERIES 79 LSET RHU in namelist NAM_PERT_PRE 29 LSETTADJ in namelist NAM PARAM CONVECTn 68 LSHAL in namelist NAM PARAM CONVECTn 68 LSHIFT in namelist NAM REAL CONF 157 LSIG CONV in namelist NAM TURB
119. NAM_VER_GRID 31 ZSTRGRD in namelist NAM_VER_GRID 32 ZSTRTOP in namelist NAM_VER_GRID 32 ZZMAX_STRGRD in namelist NAM VER GRID 32 158 159 159 159 159 INDEX
120. NF 52 RRC 87 RRI 89 RRV 86 RTH 85 RRH 91 RRR 88 RTH 85 RRC 87 RRI 89 RTH 85 RRC 87 RRI 89 RRV 86 RSV 91 RTH 85 RU 83 RV 84 RW 84 in namelist NAM PARAM CONVECTn 67 in namelist NAM VPROFn PRE 34 in namelist NAM DIMn 25 in namelist NAM MESONH DOM 147 in namelist NAM BU RRI 89 238 NIMLTRC in namelist NAM_BU_RRC 87 NIMLTTH in namelist NAM_BU_RTH 85 NITR in namelist NAM_DYNn_PRE 26 in namelist NAM_DYNn 63 NIZS in namelist NAM_GRIDH_PRE 27 NJLOC in namelist NAM_VPROFn_PRE 34 NJMAX in namelist NAM_DIMn_PRE 25 in namelist NAM MESONH DOM 147 NJZS in namelist NAM_GRIDH_PRE 27 NK in namelist NAM HURR CONF 160 NKMAX in namelist NAM VER GRID 31 158 NKWH in namelist NAM_PERT_PRE 29 NLITER in namelist NAM ADVn 61 NMODE DST in namelist NAM DUST 53 NMODEL in namelist _ 52 NMODEL_CLOUD in namelist NAM_TURB_CLOUD 60 NNEGARG in namelist NAM_BU_RRG 90 NNEGARH in namelist NAM_BU_RRH 91 NNEGARI in namelist NAM_BU_RRI 89 NNEGARS in namelist NAM_BU_RRS 90 NNEGATH in namelist NAM_BU_RTH 85 NNESTRC in namelist NAM_BU_RRC 87 NNESTRG in namelist NAM_BU_RRG 90 NNESTRH in namelist NAM_BU_RRH 91 NNESTRI in namelist NAM_BU_RRI 89 NNESTRR NNESTRS NNESTRV NNESTSV NNESTTH NNESTU NNESTV NNESTW NNUDRV NNUDTH NNUDU NNUDV NNUDW NPHIL NPREFTH NPRESU NPRESV NPRESW
121. NONE NBUMOD 1 XBULEN 300 NBUKL NBUJL 3 NBUJH 18 LBU ICP T LBU_JCP T NBUMASK 2 BU RU LBU_RU F NASSEU 1 NADVXU 1 NADVYU 1 NADVZ NHTURBU 1 NVTURBU 1 NPRESU 1 BU RV LBU_RV F NASSEV 1 NADVXV 0 NADVYV 0 NADVZ NHTURBV 0 NVTURBV 0 NPRESV O amp NAM BU RW LBU_RW F NASSEW 1 NADVXW 0 NADVYW 0 NADVZ NRELW 0 NHTURBW 0 NVTURBW 0 NPRESW O BU RTH LBU_RTH F NASSETH 1 NADVXTH 1 NADVYTH 1 NRADTH 0 NDCONVTH O NHTURBTH 1 NVTURBTH 1 NREVATH NACCTH 0 NCFRZTH 0 NWETGTH 0 NDRYGTH 0 NGMLTTH O BU LBU_RTKE F NASSETKE 1 NADVXTKE 0 NADVYTKE NDISSTKE 0 NTRTKE 0 BU RRV LBU_RRV F NASSERV 1 NADVXRV 0 NADVYRV 0 NHTURBRV 0 NVTURBRV 0 NREVARV 0 NCONDRV 0 NHENURV amp NAM BU RRC LBU_RRC F NASSERC 1 NADVXRC O NADVYRC O NVTURBRC 0 NACCRRC 0 NAUTORC 0 NCONDRC 0 NRIMRC O BU RRR LBU_RRR F NASSERR 1 NADVXRR 0 NADVYRR 0 NREVARR 0 NSEDIRR 0 NACCRR 0 NCFRZRR 0 NWETGRR O BU RRI LBU_RRI NASSERI 1 NADVXRI 0 NADVYRI 0 F CCHEM INPUT FILE MNHC input LIT SUBSTEPS 1 1 CEQNSYS DUR 0 OT 17000 1 XWAY O E 0 7 2 18000 8 21600 8 25200 8 28800 96 2 NBUKH 46 LBU XBUWRI 300 NE U 1 NFRCU 0 NCURVU 1
122. OCAL HOST REMOTE HOST 52 Thr zio 91 bc Copy the use library in the TEMP U b directory if aready exists T N R E i 2 SIMUL P i TEMP prepsourcerc S F inprepsource lib mastern a pure d P J C I I E i i Eod i t 5 x z e rs 24 L p ees o i m Ww takes ouf the Fortran G F sources fsplits b I X cU RE opor EE i a l Figure 7 1 different locations of the informations 7 3 THE PREPSOURCE PROCEDURE 203 BIBMASTER something a BIBBUGFIX 0 and BIBUSER anotherthing a This is the typical user case in which the new object code is stored in BIBUSER and missing object code will be extracted from BIBMASTER without bugfix BIBMASTER something a BIBBUGFIX bugfix_name a and BIBUSER anotherthing a In this case which should be a most common way of using the meso nh procedures the BIBUSER library is used containing the modified source files objects with bugfix BIBMASTER 0 BIBBUGFIX 0 and BIBUSER 0 Here the job will run with all the object code that is generated during the prepsource step This code is then stored in a temporary library called templib a that will not be saved This is a precarious test configuration all that will be left from the job is the output listing BIBMASTER 0 BIBBUGFIX 0 and BIBUSER something a Th
123. OON time length of trajectories computation centered on CURrent time s e XSTEP_AIRCRAFT_BALLOON minimum time step for trajectories computation s e XLAT_BALLOON initial latitudes of the balloons e XLON_BALLOON initial longitudes of the balloons e XALT_BALLOON initial altitudes of the balloons m 2 Namelist NAM_DIAG_BLANK Fortran type default value XDUMMY_DIAG array real 20 0 NDUMMY_DIAG array integer 20 0 LDUMMY DIAG array logical 20 TRUE CDUMMY DIAG array 80 characters 20 identical use than NAM BLANK see chapter 4 Add USE MODD DIAG BLANK in a diag subroutine to use any of these variables 3 Namelist NAM DIAG FILE default value YINIFILE array of character len 28 24 YSUFFIX character len 5 DIAG e YINIFILE name of the input FM files e YSUFFIX suffix appended to input file name to form output file name 4 Namelist NAM STO FILE controls trajectories computation only read if LTRAJ TRUE in NAM DIAG 180 CHAPTER 6 COMPUTE DIAGNOSTICS AFTER A MESO NH SIMULATION Fortran type default value CFILES array of character len 28 100 NSTART SUPP array of integer 100 NUNDEF e CFILES name of all the input FM files taken into account to compute trajectories They must be in inverse chronological order and correspond to a reinitialisation of Lagrangian tracers see chapter 4 e NSTART_SUPP extra origins for trajectories computation In the secon
124. P REAL CASE PREP_PGD PREP_NEST_PGD ZOMM PGD MODEL DIAG SPAWNING the namelist file name 6NAMELIST FILE and the variables to get in SLISTGET are then automatically initialized see tab 2 1 We can distinguish six steps which will be executed on the remote host see figure 2 2 1 it performs the load step if the binary libraries of object files have been modified and saves the executable ABS MAINPROG exe in workdir or it only copies it from workdir to TEMP 2 the input files read in the namelist file are search and copy to TEMP execdir If the files are got from the storage file machine they are also copied on workdir INDIR directory to prevent next transfer The next time the file will be got from this directory if it is still here 3 it executes ABSSMAINPROG exe i e the fortran code in TEMP execdir 4 it saves the output files on remote host workdir OUTDIR or HOME OUTDIR or on the storage file machine SHOME OUTDIR system In the latter case it also copies the output files on the workdir OUTDIR directory to prevent next transfers 5 it verifies if output files are still present and were not yet saved If necessary it saves them on remote host or the storage file machine 6 it sends to the local host a listing including all the echoes of the UNIX orders and a copy of the outputs directly generated by the Fortran program Again control variables are needed These parameters are set up in a fi
125. RACTER LEN INTENT IN HLUOUT Name of the output listing TYPE DATE TIME INTENT OUT TPDTCUR Current date and time Ix 0 2 declarations of local variables REAL ZCFL CFL number REAL ZWIND MAX maximum value of the wind REAL DIMENSION SIZE PUT 1 SIZE PUT 2 SIZE PUT 3 ZWIND MOD modulus of the horizontal wind CHARACTER LEN 40 YTITLE Title for date print INTEGER ILUOUT IRESP Logical unit number associated with HLUOUT and return code INTEGER IIB IJB IKB IIE IJE IKE index values for the inner mass points 1 206 CHAPTER 7 MODIFY THE FORTRAN SOURCES IIE SIZE PUT 1 JPHEXT IJB 1 JPHEXT IJE SIZE PUT 2 JPHEXT IKB 1 JPVEXT IKE SIZE PUT 3 JPVEXT IF NOT 12 THEN ZWIND MOD SQRT MXF PUT 2 MYF PVT 2 ZWIND MAX MAXVAL ZWIND MOD IIB IIE IJB IJE IKB IKE ELSE ZWIND MAX MAXVAL PUT IIB IIE IJB IJE IKB IKE END IF See soe So ae Se Se ee Se ee ee Se eee eS eS Se ee Se SS ee eee So ee ee ee I 2 COMPUTES THE CFL NUMBER IF NOT L2D THEN ZCFL ZWIND_MAX PTSTEP MAX MAXVAL PDXHAT MAXVAL PDYHAT ELSE ZCFL ZWIND_MAX PTSTEP MAXVAL PDXHAT END IF See ee ee a ee ee ee ee ee ee ee ee ee See ee I 3 PRINTS THE WARNING FOR CFL gt 1 IF ZCFL gt 1 THEN CALL FMLOOK_11 HLUOUT HLUOUT ILUOUT IRESP YTITLE WARNING FROM CFL CALL SM PRINT TIME TPDTCUR HLUOUT YTITLE WR
126. RFORM A MESONH SIMULATION If the coupling files are given by CCPLFILE D2 amp CCPLFILE 2 F gt t CCPLFILE 3 A2 gt t CCPLFILE 4 A5 t4 CCPLFILE 5 2 NONE gt CCPLFILE 8 gt then the instants must satisfy tsegment lt lt lt t3 lt t4 If it is not the case the program stops The case for which the coupling fields are not time dependent does not require any coupling file because these coupling fields are read in the initial MESONH file of model 1 as the Larger scale fields LSUM LSVM LSWM LSTHM LSRVM More details can be found in the scientific documentation of the model 4 2 16 Namelist NAM NUDGINGn nudging of model n Fortran type default value LNUDGING logical FALSE XTNUDGING real 21600 It contains the parameters needed for nudging of U V W TH Rv fields of model n towards Large Scale values They are included in the declarative module MODD NUDGINGn e LNUDGING switch to activate nudging for model n e XTNUDGING time scale for nudging towards Large Scale values 4 2 17 Namelist NAM PARAMn parameterizations names of model n Fortran type default value CTURB 4 characters CRAD 4 characters CCLOUD 4 characters CDCONV 4 characters It contains the types of the different parameterizations used by the model n They are included in the declarative module MODD PARAMn e CTURB gives the type of turbu
127. Sky fraction and Random Overlap for Effective Zenithal Angl DEFAULT VALUE This option is well adapted to multi layer clouds e NOVLP 1 Maximum overlap for Clear Sky fraction and Random Overlap for Effective Zenithal Angl This option is well adapted in the absence of multi layer clouds e NOVLP 8 Maximum Random overlap for Clear Sky fraction and Effective Zenithal Angl It corresponds to the previous configurations before masdev4 7 74 CHAPTER 4 PERFORM A MESONH SIMULATION 4 2 20 Namelist NAM_TURBn turbulence parameters for model n Fortran type default value XIMPL real T CTURBLEN 4 characters BL89 CTURBDIM 4 characters 1DIM LTURB_FLX logical FALSE LTURB DIAG logical FALSE LSUBG COND logical FALSE LSUBG_AUCYV logical FALSE LSIGMAS logical TRUE LSIG_CONV logical FALSE LRMCO1 logical FALSE CTOM 4 characters NONE It contains the characteristics of the turbulence scheme used by the model n They are included in the declarative module MODD_TURBn e XIMPL degree of implicitness of the vertical part of the turbulence scheme XIMPL 0 5 corresponds to the Cranck Nicholson scheme for the vertical turbulent diffusion and 0 to a purely explicit scheme e CTURBDIM turbulence dimensionnality CTURBDIM 1DIM Only the vertical turbulent fluxes are taken into account This has to be done for relatively large horizontal meshes CTURBDIM 3DIM the turbulent fluxes
128. TER 5 INITIALIZATION OF MESO NH FOR REAL CASES For a grid nesting simulation the hurricane filtering is first applied for the outer domain dad model with the program PREP REAL CASE The filtered fields are then horizon tally interpolated for inner domains with the program SPAWNING see section 5 8 Then for each inner domain a vortex bogussing is added with the program PREP REAL CASE LFILTERING logical FALSE CFILTERING character LEN 5 UVT NK integer 50 XLAMBDA real 0 2 XLATGUESS real XUNDEF XLONGUESS real XUNDEF XBOXWIND real XUNDEF XRADGUESS real XUNDEF NPHIL integer 24 NDIAG FILT integer 1 LBOGUSSING logical FALSE XLATBOG real XUNDEF XLONBOG real XUNDEF XVTMAXSURF real XUNDEF XRADWINDSURE real XUNDEF CDADATMFILE character LEN 28 CDADBOGFILE character LEN 28 gt LFILTERING to switch on the filtering of the fields U V T reduced Ps of the atmospheric file logical CFILTERING to choose the list of the fields to be filtered U V T reduced Ps U V T are filtered default UVTP U V T and reduced PS are filtered NK number of points of the half width of the window in which the Barnes filter is applied to compute low pass component of a given field XLAMBDA a coefficient in the exponential weighting function of the Barnes filter XLATGUESS latitude of the guessed position of the cyclone center XLONGUESS longitude of the guessed position of the cyclone cent
129. TER 8 THE MESONH FILES e To PREpare an initial Meso NH file for an IDEAlized case study gt file PRE_IDEA1 nam To PREpare an initial Meso NH file for an REAL case study gt file REALI nam e To SPAWN an Meso NH file into another one with better horizontal resolution gt file SPAWNI nam e To EXecute a simulation SEGment for the nt model gt file EXSEGn nam e To compute DIAGnostics after a simulation gt file DIAG1 nam Because the grid nesting technic requires the simultanous presence of multiple models in the computer memory free parameters must be fixed for every model This is performed by associating one namelist file per model this explains why the namelist are suffixed by a number 1 or n just above The different parameters present in these files are all given in this book chapters 3 to 6 and more details on the description of a given parameter are present in the code itself 8 2 The Meso NH files Meso NH FM file is a set of 2 files sticked together via the cpio UNIX command the 2 parts are e a descriptive part des containing informations about the file generation and its de scription in ASCII characters e a binary part 1fi where the fields are stored The structure of this file is a direct access type file written and read by routines developped in M t o France Fischer 1994 based on LFI routines Clochard 1989 which can be used on a lot of different computers This type of
130. TERACTIVE e Now you may create a file which contains the following namelists FILE EXSEG1 nam for the previous example of prep ideal case 132 CHAPTER 4 PERFORM A MESONH SIMULATION amp NAM LUNITn CINIFILE HYD2D amp NAM CONFn LUSERV T DYNn 60 LITRADJ T LHORELAX UVWTH T LVE RELAX T NRIMX 5 NRIMY 3 XRIMKMAX 00166 XT4DIFF 1500 PARAMn CTURB TKEL CRAD NONE CCLOUD NONE amp NAM_TURBn XIMPL 1 CTURBLEN BL89 CTURBDIM 1DIM LTURB_DIAG T LTURB_FLX T amp NAM LBCn CLBCX 2 CLBCY 2 CYCL XCPHASE 20 amp NAM_CONF CCONF START LTHINSHELL T L2D T LFLAT F NMODEL 1 CEQNSYS DUR NVERB 1 CEXP EXPER CSEG HYD2D amp NAM_DYN XSEGLEN 20000 XASSELIN 0 2 LCORIO F XALKTOP 0 005 XALZBOT 12570 LNUMDIFF T amp NAM_FMOUT XFMOUT 1 1 10000 XFMOUT 1 2 20000 amp NAM_BLANK FILE EXSEG1 nam for a real case LUNITn CINIFILE 16J36 1 12B18 001 CCPLFILE 1 16JAN_O6_MNH CONFn LUSERV T LUSERC T LUSERR LUSERI T LUSERS T LUSERG T LUSECI T DYNn XTSTEP 75 CPRESOPT RICHA NITR 8 LHORELAX_UVWTH T LHORELAX_RV T LVE_RELAX T NRIMX 5 NRIMY 5 XRIMKMAX 0 0083 XT4DIFF 5000 amp NAM ADVn CMET_ADV_SCHEME FCT2ND CSV ADV SCHEME FCT2ND PARAMn CCLOUD CTURB TKE
131. THE MESONH USER S GUIDE MASDEVA T version April 2 2008 Contents 1 2 3 Introduction 7 The MESONH procedures except prepsource 11 2 1 Presentation d Palit ke S Gu A SO a PS 11 2 1 1 What you do with the MESONH procedures 11 2 1 2 The environment variables 13 2 1 3 The input parameters of a MESONH script 15 2 2 A complete Meso NH work session 16 2 2 1 Source extraction and compilation prepsource 16 2 2 2 Running the model Preparing an initial file or Computing diagnoctic fields prepmodel MM rl we eR eh A 16 2 2 8 Send your job for execution tosupc 18 Prepare an initial MESONH file for an ideal case 21 3 1 Overview of PREP IDEAL CASE functionalities 21 3 2 The input the PREIDEAI1 namfile a 22 3 21 Namelist NAM BLANK available variables 22 3 2 2 Namelist NAM CH MNHOCn PRE init chemistry scalar variables 22 3 2 3 Namelist NAM CONF PRE configuration 23 3 2 4 Namelist NAM CONFn configuration variables for modeln 25 3 2 5 Namelist NAM DIMn contains dimensions 25 3 2 6 Namelist NAM DUST PRE init dust scalar variables 25 3 2 7 Namelist NAM DYNn pressure solver 26 3 2 8 Namelist NAM GRID PRE g
132. U WSV SBG DP M dyn prod by mean gradient BU WSV 5 prod by resolved fluctuations BU WSV SBG TR subgrid turbulent transport BU WSV SBG PRES z t n subgrid pressure correlation term ES BUOWSV SBGTP atm thermal production residual of budget of Sw gt BU WSV_ REST tn must be small 44 BUDGET OF RESOLVED SUBGRID TURBULENT QUANTITIES 125 field notation in processus dim comments ed i ie ane _ m BU WSV RES TEND zia opposite of tendency of c gt by mean flow US BU WSV RES FORC z t n advection by large scale BU WSV RES DP ztn dyn prod meas BU WSV RES TR z t n transport of resolved flux by itself BUOVSV RES PRES BU WSV_ REST v BU WSV RESSBGT 2 6 sink due to subgrid turbulence uw gt BUWSV RESCORI zaa numerical diffusion of lt ws gt BU WSV RES NUMD z t n numerical diffusion against 2Ax relaxation of lt BU WSV RES RELA sponge layer relaxation 2way nesting of lt wsy gt BU_WSV RES NEST z t n average from smaller nested models miscellaneous 2767 RES MISC zio curvature term chemistry residual of budget of lt ws gt BU WSV ztn musbezro 1 neglected in turb scheme tendency of lt w s gt w Zi us neglected advection by mean flow Ds B
133. UD 60 CTYPELOC in namelist NAM_VPROFn_PRE 33 CUVW_ADV_SCHEME in namelist NAM_ADVn 60 CZS in namelist NAM_CONF_PRE 23 G GBAL_ONLY INDEX in namelist NAM_GRID2_SPA 166 H HATMFILE in namelist NAM_FILE_NAMES 156 169 HATMFILETYPE in namelist NAM_FILE_NAMES 156 169 HCHEMFILE in namelist NAM_FILE_NAMES 157 HCHEMFILETYPE in namelist NAM_FILE_NAMES 157 HPGDFILE in namelist NAM_FILE_NAMES 157 169 I IDAD in namelist NAM PGDI 144 in namelist NAM PGD2 144 in namelist NAM PGD3 144 in namelist NAM PGDA 144 in namelist NAM PGD5 144 in namelist NAM_PGD6 144 in namelist NAM_PGD7 144 in namelist NAM PGDS 144 IDXRATIO in namelist NAM GRID2 SPA 166 IDYRATIO in namelist NAM GRID2 SPA 166 IXOR in namelist NAM GRID2 SPA 165 IXSIZE in namelist NAM GRID2 SPA 166 IYOR in namelist NAM GRID2 SPA 166 IYSIZE in namelist NAM GRID2 SPA 166 L LBOGUSSING in namelist NAM HURR CONF 161 LBOUSS in namelist NAM CONF PRE 24 LBU ICP in namelist NAM BUDGET 82 LBU_JCP in namelist NAM_BUDGET 82 LBU_KCP in namelist NAM_BUDGET 82 LBU_RRC INDEX in namelist NAM BU RRC 87 LBU RRG in namelist NAM BU RRG 90 LBU RRH in namelist NAM_BU_RRH 91 LBU RRI in namelist NAM BU RRI 89 LBU RRR in namelist NAM BU RRR 88 LBU RRS in namelist NAM BU RRS 90 LBU RRV in namelist NAM_BU_RRV 86 LBU RSV in namelist NAM BU RSV 91 LBU RTH in namelist NAM_BU_RTH 85 L
134. VVAR 3D K2 K kg kg kg kg UW VFLX VW VFLX 3D m s W VVAR 3D m s WSV FLX n 3D SVunit m s e 3d scheme turbulent fluxes UTH FLX VTH FLX UR VR_FLX 3D K m s kg kg m s TH HVAR THR HCOR R HVAR 3D K K kg kg kg kg U VAR V VAR W VAR 3D m s UV FLX UW FLX VW FLX USV FLX n 3D m s SVunit m s e Mixing length for clouds RVCI 3D rv rc 4 ri kg kg GX RVCI GY RVCI 3D x and y gradient of RVCI kg kg m GNORM RVCI 3D Horizontal norm of the gradient of RVCI kg kg m QX_RVCI QY_RVCI 3D x and y gradient of the advection of RVCI kg kg m QNORM RVCI 3D Horizontal norm of the gradient of the advection of RVCI kg kg m CEI 3D Cloud entrainment instability index kg kg m s 184 CHAPTER 6 COMPUTE DIAGNOSTICS AFTER A MESO NH SIMULATION Radiation scheme DTHRAD 3D Radiative tendency for 0 K s NRAD_3D 0 1 2 3 4 5 FLALWD DIRFLASWD SCAFLASWD DIRSRFSWD 2D radiative fluxes W m 0 CLEARCOL_TM1 EMIS 2D trace of cloud emmissivity ZENITH AZIM 2D solar zenithal angle azimuthal angle RAD DIR ALB SCA ALB 2D direct albedo scattered albedo TSRAD 2D radiative surface temperature SWF DOWN SWF_UP LWF DOWN 3D radiative fluxes W m LWF UP LWF NET SWF NET E DTRAD LW DTRAD SW 3D Radiative tendency for T K day RADSWD VIS RADSWD NIR RADLWD 2D surface radiative flux W m SWF DOWN CS SWF UP CS
135. X 27PHEXT 3 2 6 Namelist NAM DUST init dust scalar variables If you initialize aerosol passive dust during PREP IDEAL CASE use the following namelist variables default value EDUST FALSE e LDUST switch to activate initialization of passive dust 3 modes 26 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE 3 2 7 Namelist NAM DYNn pressure solver Fortran type default value CPRESOPT 5 characters CREST NITR integer 4 XRELAX real CPRESOPT gives the type of pressure solver used for the elliptic equation RICHA CRESI This equation is solved in order to ensure the anelastic constraint for the initial wind field Note that the solver is applied even for the FLAT case when the Earth spericity is taken into account NITR number of iterations used for the elliptic equation resolution solver CPRE SOPT e XRELAX relaxation factor used by the Richardson method CPRESOPT RICHA 3 2 8 Namelist NAM GRID PRE grid definition Fortran type default value XLONO XLATO XBETA XRPK XLONORI XLATORI XLONO reference longitude for conformal projection if LCARTESIAN TRUE this value can be usefull to compute local solar time XLATO reference latitude for conformal projection and cartesian plane XBETA rotation angle for conformal projection and cartesian plane XRPK cone factor for the projection not used if LCARTESIAN
136. YFILETYPE DIRECT YSAND sand_fao YSANDFILETYPE DIRECT XUNIF_RUNOFFB 0 5 amp NAM_CH_EMIS_PGD NEMIS_PGD_NBR 0 amp NAM DUMMY PGD NDUMMY_PGD_NBR 0 And for the son PGD file amp NAM PGDFILE CPGDFILE PGD_AMMA1_5km_m46_b1 amp NAM PGD SCHEMES CNATURE ISBA CSEA SEAFLX CWATER WATFLX CTOWN TEB amp NAM PGD GRID YINIFILE PGD_AMMA1_10km_m46_b1 YFILETYPE MESONH INIFILE CONF PROJ IXOR 101 IYOR 21 IXSIZE 180 IYSIZE 150 IDXRATIO 2 IDYRATIO 2 amp NAM_COVER YCOVER ecoclimats_v2 YFILETYPE DIRECT LRM_TOWN FALSE amp NAM_ZS YZS gtopo30 YFILETYPE DIRECT COROGTYPE AVG XENV 0 NZSFILTER 1 amp NAM_ISBA NPATCH 12 CISBA 3 L CPHOTO AGS NGROUND_LAYER 3 YCLAY clay_fao YCLAYFILETYPE DIRECT YSAND sand_fao YSANDFILETYPE DIRECT XUNIF_RUNOFFB 0 5 amp NAM_CH_EMIS_PGD NEMIS_PGD_NBR 0 amp NAM DUMMY PGD NDUMMY_PGD_NBR 0 148 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES 5 3 Modification of PGD files for grid nesting PREP NEST PGD In order to run models with the gridnesting technique a condition on the orography must be satisfied In the following if file 2 is completely included and therefore in interaction during the run with file 1 file 2 will be called the SON file and file 1 the DAD file In the following the DAD file number must be smaller th
137. Z case and the function F Y is a simple function of Y which must be adapted by modifying its definition directly in the routine FUNUY The default function is FUNUY 1 9 00 cosh zwidth 3 2 THE INPUT THE PRE_IDEA1 NAM FILE 33 Y Z U G Y Z The function G must also be adapted by modifying its definition directly in the routine FUNUYZ The default function is 1 2 2 cosh dh T Notice that in this case the U z values given by the profile are not used FUNUYZ f z e CFUNV String of 3 characters describing the type of function which gives the y com ponent of the wind Possible configurations are listed below ZZZ V V z The V z values are taken from the Radio Sounding or analitycal profile given in the free formatted part of the PRE IDEA nam file X Z V F X V Z The V z values are build in the same way as the ZZZ case and the function F X is a simple function of X which must be adapted by modifying its definition directly in the routine FUNVX The default function is 1 cosh 35 X Z V G X Z The function G must also be adapted by modifying its definition directly in the routine FUNVXZ The default function is FUNVX amp 1 SI cR cosh Geist Notice that in this case the V z values given by the profile are not used FUNVXZ amp z e CTYPELOC Type of informations used
138. acters integer 1000 integers real real integer integer integer logical 1 character 1 character integer real real real integer real real real SIS 0 1000 0 0 001 0 1 4 2 25 Simulation of the fly of balloons or aircraft in the model fields In order to compare the model outputs to airborne observations and measurements it can be interesting to simulate the movement of a balloon or an aircraft during the model run in any one model for gridnesting runs A balloon is launch at a given location and either for a particular density iso density balloon a particular volume constant volume balloon or ascent speed radio sounding For iso density balloons initial altitude or pressure is asked A balloon is advected by the wind of the model It can crash For an aircraft the flight legs must be given by the user location and duration All the prognostic fields zonal and meridien wind from U and V components vertical velocity potential temperature pression mixing ratios tke radiative surface temperature are recorded on the trajectory of the balloon or the aircraft as well as the trajectory itself position in X Y and Z directions and orography All records are in the diachronic file 000 Up to 9 balloons and aircraft are possible The specification of the characteristics of flights are not given in a namelist but directly in Fortran routines e ini balloon f90 for balloons e ini aircraft f9
139. al domain and the horizontal grid e the physiographic fields 5 7 3 The atmospheric file Both GRIB and FM file are self explanatory The physiographic data stored in it will not be saved on the output MESONH file 5 7 4 The chemical file optional Both GRIB and FM file are self explanatory If the atmospheric file is a GRIB file the chemical species can be read in another file than the atmospheric one 5 7 5 The file REAL1 nam This file contains namelists with the directives to run PREP REAL CASE The namelists con tain the names of the files and the definition of the vertical grid T he order of namelist is free and unset namelists can be ommited The file can also contain a free formatted part after the vertical grid definition namelist where are given the vertical levels if this option is chosen 1 Namelist NAM FILE NAMES contains file names HATMFILE character LEN 28 gt HATMFILETYPE character LEN 6 MESONH HPGDFILE character LEN 28 HCHEMFILETYPE character LEN 6 MESONH CINIFILE character LEN 28 INIFILE HCHEMFILE character LEN 28 e HATMFILE name of the atmospheric file e HATMFILETYPE type of the atmospheric file GRIBEX MESONH 5 7 VERTICAL INTERPOLATIONS PREP REAL CASE 161 e HPGDFILE name of the Physiographic Data File e HCHEMFILE optional name of the file containing the chemical species if they are not in
140. amelist in namelist CLBCY in namelist in namelist CLW in namelist NAM CONF 53 NAM HURR CONF 160 NAM VPROFn PRE 32 NAM VPROFn PRE 33 NAM CONF PRE 23 NAM LUNIT2 SPA 166 NAM LUNITn 28 65 NAM CONF 52 NAM_LBCn_PRE 28 NAM LBCn 64 NAM LBOCn PRE 28 NAM LBCn 64 NAM PARAM RADn 69 CMET_ADV_SCHEME in namelist CMINERAL in namelist CNUCLEATION in namelist COPILW in namelist COPISW in namelist COPWLW in namelist COPWSW in namelist CORGANIC in namelist 229 NAM ADVn 61 NAM CH ORILAM 77 NAM CH ORILAM 77 NAM PARAM RADn 70 NAM PARAM RADn 70 NAM PARAM RADn 70 NAM PARAM RADn 70 NAM CH ORILAM 77 NAM FILE NAMES 157 169 230 CPERT KIND in namelist NAM_PERT_PRE 29 CPGD FILE in namelist NAM_REAL_PGD 30 CPGD FILE in namelist NAM_REAL_PGD 30 CPGDFILE in namelist NAM_PGDFILE 141 146 CPRESOPT in namelist NAM_DYNn_PRE 26 in namelist NAM_DYNn 63 in namelist NAM REAL CONF 157 CPRISTINE ICE in namelist NAM PARAM ICE 59 CRAD in namelist NAM PARAMn 66 CRELAX HEIGHT in namelist NAM_FRC 56 CRGUNIT in namelist NAM AERO CONF 162 in namelist NAM CH ORILAM 76 CSEG in namelist NAM_CONF 53 CSPLIT in namelist NAM CONF 52 CSURF in namelist NAM_GRn_PRE 23 28 CSV_ADV_SCHEME in namelist NAM_ADVn 61 CTOM in namelist NAM_TURBn 72 CTURB in namelist NAM_PARAMn 66 CTURBDIM in namelist NAM_TURBn 71 CTURBLEN in namelist NAM_TURBn 72 CTURBLEN_CLOUD in namelist NAM_TURB_CLO
141. an any of its SON number The condition on the orography is the mean of orography for a SON file in the domain corresponding to the grid mesh of its DAD file must be equal to the orography of the DAD file in this mesh Such a condition is not automatically satisfied when using enhanced orographies program PREP_NEST_PGD performs post treatments on the orographies of up to 8 PGD files that will be used to create initialization files for a gridnested run It modifies the orography of a DAD from the mean of the orography of its several SON s PREP_NEST_PGD is run with the MESO NH procedure prepmodel In the file prepmodelrc the input host directories and login control variables refer to the input PGD files The other control variables to initialize specifically in this file are e MAINPROG PREP_NEST_PGD e NAMELISTFILE default e LISTGET default The namelist file NEST PGDI nam contains the order of namelists is free and unset namelists can be ommited 1 Namelists NAM PGDN where N goes from 1 to 8 e YPGDN name of the PGD file N e IDAD number of the DAD file of file The DAD file number must be smaller than N 2 Namelist NAM NEST PGD e YNEST string of 2 characters to be added to the PGD file names to define the corresponding output PGD file names The input file YPGDN will be modified into file YPGDN nest YNEST Example of namelist PRE_NEST_PGD1 nam 5 3 MODIFICATION OF PGD FILES FOR GRI
142. and W at KCLA are stored e NKLOW NKUP two K levels x t series of mean W between KLOW and KUP and mean Rc between the ground and KUP are stored e NKMID a K level x t serie of Rv at KMID is stored e NBJSLICE number of y slices for x t serie e NJSLICEL NJSLICEH lower and higher index along y axe of the y slices e NFREQSERIES Time frequency of diagnostic writing 84 CHAPTER 4 PERFORM A MESONH SIMULATION 4 2 29 Namelist BUDGET budget box description Fortran type default value CBUTYPE 4 characters NBUMOD integer XBULEN real NBUKL integer NBUKH integer LBU KCP logical XBUWRI real NBUIL integer NBUIH integer NBUJL integer NBUJH integer LBU ICP logical LBU_JCP logical NBU_MASK integer It contains the description of the box in which the budget are performed This box is always built with a subset of points of the simulation box e CBUTYPE type of box used to compute the budget CART a cartesian box defined by the lowest and highest values of the indices in the 3 directions in the MESONH grid defined in the following MASK several areas described by horizontal masks are selected according to criteria evaluated at each model timestep budget computations are realized at the selected verticals for each criteria criteria are to be defined in the routine set mask f90 e NBUMOD number of the model in which the budget are performed Only one m
143. another file y or n default y 8 gt enter the root directory from which source files will be extracted 8 gt default 8 gt ist users come fischer sources modif 8 2nd users come fischer mnh masterO 8 3rd mesonh sources 8 9 gt please enter arguments for nh takeout file name version name VID SOIDVUL HN OS r4 ory oo1nos Soy py amS nh takeout source f90 MASTER 1 1 local file nh unregister source f90 MASTER local file nh create source f90 MASTER 1 4 nh extract source f90 MASTER local file nh register source f90 MASTER 1 2 nh extract source f90 MASTER local file nh register source f90 MASTER pm 295 ae nh extract source f90 MASTER 1 1 local file nh unextract source f90 MASTER LdIWOS AHL AO W IdIWWVXMH cv EGG 224 APPENDIX AN ILLUSTRATIVE EXAMPLE OF USE OF PROCEDURES init f90 MASTERO 10 init f90 already exists on users come fischer mnh masterO 10 gt do you want to overwrite it y or n default n n 11 gt do you want another file y or n default y n 12 gt enter the global parameters for compilation on remote machine makefile 12 gt compilation options default 13 gt end of silent interactive prepsource loop 13 gt a new file inprepsource new is now available 13 gt for future work with prepsource in SILENTINTERACTIVE 13 mod dei n f90
144. aphic fields e Ecoclimap file describing the type of cover of the surface at 30 of resolution The file is ecolimap v2 e A file containing the orography The resolution of the MESO NH team file is 30 on the world This allows to compute the model orography and some subgrid scale orographic characteristics The file is gtopo30 e file with the clay fraction of the near surface soil The resolution of the file of the MESO NH team is 5 on the world The file is clay fao e file with the sand fraction of the near surface soil The resolution of the file of the MESO NH team is 5 on the world The file is sand fao 5 2 CREATION OF MESO NH PHYSIOGRAPHIC DATA FILE PREP_PGD 145 5 2 3 The input PREP PGDI nam file The file PRE_PGD1 nam is needed containing several namelists The order of namelists is free and unset namelists can be ommited 1 Namelist NAM_PGDFILE contains file names default value CPGDFILE character LEN 28 e CPGDFILE name of the output Physiographic Data File 2 Namelists of the externalized surface for PREP_PGD As indicated above the further definition of the surface parameters are not done by MESONH itself but by the externalized surface included in it So you are invited to refer to the documentation of the surface You must fill the following namelists 5 2 4 NAM PGD SCHEMES NAM PGD GRID NAM CONF PROJ NAM CONF PROJ GRID if you define the grid co
145. arameterized values in ice subgrid condensation scheme LSIG CONV Switch for computing Sigma s due to convection in ice subgrid condensation scheme e LRMCOI Switch for computing separate mixing and dissipative length in the SBL ac cording to Redelsperger Mahe and Carlotti 2001 e CTOM Consideration of Third Order Moments CTOM NONE No Third Order moments CTOM TMO06 Parameterization of Third Order moments of heat fluxes for dry CBL according to Tomas and Masson 2006 4 2 21 SURFACE SCHEMES namelists of the externalized surface The further definition of the surface parameters are not done by MESONH itself but by the externalized surface included in it So you are invited to refer to the documentation of the surface To summarize the following namelists must be defined version 1 of the externalized sur face physical schemes for sea and for vegetation e NAM SEAFLUXn e NAM ISBAn Chemical schemes for each surface e NAM CH CONTROLn 76 CHAPTER 4 e NAM_CH_SURFn e NAM_CH_SEAFLUXn e NAM_CH_WATFLUXn e NAM CH ISBAn e NAM_CH_TEBn Online computation of diagnostics e NAM DIAG SURF ATMn e NAM DIAG SURFn e NAM DIAG ISBAn e NAM DIAG TEBn PERFORM MESONH SIMULATION 4 2 22 CHEMISTRY scheme Namelist NAM CH MNHOCn control of MNHC Fortran type default valne LUSECHEM LCH INIT FIELD LCH_SURFACE_FLUX LCH CONV SCAV LCH EXPLICIT SCAV LCH CONV LINOX
146. ard information to the father also for 2D fields Surface precipitation and SW radiative fluxes that are used by the surface 4 2 8 Namelist NAM PARAM C2R2 control variable of the 2 moment warm microphysical schemes C2R2 and KHKO Fortran type default value HPARAM CCN HINI_CCN HTYPE_CCN XCHEN XKHEN XMUHEN XBETAHEN XCONC_CCN XR MEAN CCN XLOGSIG_CCN XFSOLUB_CCN XACTEMP_CCN XALPHAC XNUC XALPHAR XNUR LRAIN LSEDC LACTIT character LEN 3 character LEN 3 character LEN 1 real real real real real real real real real real real real real boolean boolean boolean 60 CHAPTER 4 PERFORM A MESONH SIMULATION It contains the control parameters used when calling the C2R2 warm microphysical scheme They are in the declarative module MODD PARAM C2R2 e HPARAM Acronym of CCN activation parameterization to use CPB or TWO and TWO need only to prescribe the XCHEN and XKHEN parameters TWO refers to the classical activation spectrum of Twomey in the form Nocn s TFH includes some improvements brought by Feingold and Heymsfield JAS 1992 to the original activation spectrum of Twomey CPB refers to an activation spectrum in the form defined in Cohard et al JAS 1998 with Necn s Cs F u E 1 s where F is the hypergeometric function and C k 8 four adjustable coefficients e CC
147. are computed this is necessary for small horizontal meshes meso y scales or LES e CTURBLEN type of turbulent mixing length CTURBLEN If CTURBDIM 3DIM the cubic root of the grid volum is used in 3D simulations and the squared root of the volum in 2D simulations If CTURBB 1DIM we take Az in simulation of any dimensionality This length is always limited by amp z near the ground CTURBLEN BL89 The mixing length is computed according to the Bougeault and Lacarr re scheme refer to the scientific documentation CTURBLEN DEAR the mixing length is given by the mesh size depending on the model dimensionality this length is limited with the ground distance and also by the Deardorff mixing length pertinent in the stable cases e LTURB_FLX logical switch to compute and store all the turbulent fluxes on every output synchronous files 42 THE INPUT EXSEGS N NAM FILE 75 LTURB DIAG logical switch to store diagnostic quantities related to the turbulent scheme on every output synchronous files mesh length Prandtl number Schmidt number sources of TKE LSUBG COND switch to activate the subgrid condensation scheme refer to the scientific documentation for more details e LSUBG_AUCYV switch to activate the subgrid autoconversion scheme if LSUBG COND is set to TRUE and only with the mixed phase for the moment LSIGMAS Switch for using Sigma s from turbulence scheme instead p
148. are computed and stored in the diachronic group BU RT2 All comments made for the total equation are valid here ADVM ADVR DPM eo gt Ow lt r gt lt gt lt gt lt 2 gt dnm 2 lt gt Oxo E gt TEA ur gt T dO Dr t WY DISS DPR TR ADV DP 2 n lt T gt gt lt U gt o R gt 2 lt tah gt 9 3 lt uri gt 2 lt gt 9 TR SBGT field notation in processus dim comments diac file name 2 lt wri gt lt r gt BU RT2 SBGDPM dyn prod by mean gradient lt 2 gt BU RT2 SBG DP R dyn prod by resolved fluctuations 2 3 RPTE 2 lt gt BU RT2 SBG DISS dissipation residual of budget of 2 wi BUR SBG RESI at 44 BUDGET OF RESOLVED SUBGRID TURBULENT QUANTITIES 123 field notation in processus dim comments diac file name BU RI2 RES TEND opposite of tendency of lt 7 gt 7 lt w gt 2 lt gt BU RT2 RES ADV zt advection by mean flow Woredz lt gt BU RT2 RES FORC z t advection by large scale E ELI _ by mean gradients 1 resolved variance to subgrid turbulence numerical diffusion of lt gt BU RT2 RES NUMD 7 4 numerical diffusion relaxation of lt f BU RT2 RES RELA zt sponge layer relax
149. are ready to use them for the execution of one or several Meso NH main programs To do this just add in the file prepmodelrcthe name of your binary library present on the remote host in the control parameter BIBUSER For the case of the previous example file prepmodelrc looks like user binary library BIBUSER modif a reference binary library BIBMASTER masdev4 2 a reference bugfix binary library BIBBUGFIX bugil a Note that the name of the User s library created on the remote host is actually modif m4 2 a But it is not necessery to indicate the suffix linked to the Master s binary name added by prepsource the prepmodel procedure will do the same operation to retrieve the right name of the User s library name 212 CHAPTER 7 MODIFY THE FORTRAN SOURCES Chapter 8 The MESONH files 8 1 F90 namelists the informations required to perform a given step of a numerical experiment are provided by different files including NAMELIST set Thus the Meso NH user can change the value of the parameters without any compilation and therefore save computer time These files provide the way for the Meso NH user to interact with the numerical code and finally they contain the identification cards of the different steps of the numerical experiment These NAMELISTs are Fortran 90 NAMELISTs which obey to strict writing rules Metcalf and Reid 1993 no comment is allowed inside the namelists no empty namelist can be written
150. ariables are needed for MARS data other than Operational Data ERA40 and ERAI5 example for CLASS RD for data from Research Department EXPVER experience version MARS parameter DATEGAP gap to add to DATE written as YYMMDD The date DATE DATEGAP allows to determine the list of parameters to extract accord ing the version of the IFS model 0 means that the model version is the operational one at the date DATE OUTFILE name of the GRIB output file When several dates are requested the number of output file is NBLOOP 1 The name is set to ecmwf CLASS DATE TIME if TYPE AN AN ecmwf CLASS TYPE DATE TIME STEP if TYPE FC FC FLAGTO automatic submit of the script to ecgate yes no SUBMIT_NEXTJOBS name of one or several jobs to submit to ecgate at the end of this job works only if extractecmwf is executed on ecgate 5 6 EXTRACTION OF M T O FRANCE FILES EXTRACTARPEGE 155 5 6 Extraction of M t o France files extractarpege extractarpege performs the extraction of one GRIB file at M t o France center in Toulouse with fields on stretched lat lon grid for ARPEGE products on Lambert conformal projection for ALADIN forecast or on regular lat lon grid for MOCAGE outputs Remark The extraction is executed on the super computer at M t o France This allows for the extracted file to be a standard GRIB file It can be handled with C routines such as in library pbio to make the interface with fortran programs To tre
151. at its information one needs the library gribex Two main steps are performed in the extractarpege job execution 1 the post processing package FULL POS is first called to extract altitude and surface fields on the model configuration for ARPEGE and ALADIN products or a post processing program written especially in the case of MOCAGE outputs to gather surface and altitude fields on one the 3 Mocage domains 2 a tool writing the extracted fields into GRIB format is then run one is used for ARPEGE and MOCAGE products another for ALADIN forecast extractarpege has control variables These are given in the file extractarpegerc or interactively They are e MODEL name of the model ARPEGE ARPEGE Tropiques ALADIN France ALADIN Runion or MOCAGE arpifs tropic aladin reunion mocage e TYPE for ARPEGE and ALADIN for operational archive type of altitude and surface fields AN non initialized analysis IA initialized analysis FG first guess short forecast from previous initialized analysis to smooth the fields FC forecast for other archive something else for MOCAGE type of domain 095 EURO11 GLOB22 RSE088 FR EU GL SE e DATE date written as YYYYMMDD Details can be found in the FULL POS users guide by Ryad El Khatib CNRM GMAP Details about MOCAGE model can be found from V H Peuch CNRM GMGEC ERAM 156 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES TIME validity time HH
152. ation 2way nesting of lt 7 gt BU RT2 RES NEST z average from T Ls eater netea nodes miscellaneous BU_RT2 RES MISC z t ref pressure term LS residual of budget of lt A gt BURT RES RESI at must bezero 5 lt r gt neglected opposite of neglected in turb scheme tendency of lt r gt lt w gt 2 lt gt neglected advection by mean flow E lager gt neglected advection by resolved flow 124 CHAPTER 4 PERFORM A MESONH SIMULATION 4 4 7 Budget of total scalar flux terms of the equation of Z QS gt lt w s gt are computed and stored in the diachronic group BU WSV All comments made for the total Tke equation are valid here ADVM ADVR DPM 2 lt ws gt lt Ua gt ws Ws uws QE ur REUS Oza Ota 5 Ow uls r fis gt pe gt lt ul wsi gt Oza To lt p gt Oxo TP DPR PRES TR ADV PRES DP M c d lt eae EN oss cuu Pe A 2 lt 08 gt lt lt 108 lt gt lt 458 2 x Sg lt p gt E 5 Oz a i Os lt 88 0 gt lt Ws gt lt wu sil gt lt Sy uw gt a La SBGT field notation in processus dim comments diac file name lt wW gt lt 8 gt B
153. between the PREP IDEAL CASE program and the user is made through the PRE_IDEA1 nam file The degrees of freedom are collected in a set of namelists read by this program 21 22 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE 3 2 The input the PRE_IDEA1 nam file It is made of two parts e A namelist part with directives for the preparation of an idealized case always present The order of namelists is free and unset namelists can be ommited e free formatted part describing a vertical profile of n layers of constant moist Brunt Vaisala frequency or a radiosounding and sometimes the explicit list of the heights of the vertical levels This part can be present or absent in the other cases To initialize a simulation with a radiosounding and real terrain conditions it is necessary to perform the PREP PGD program see next chapter to create a MESO NH physiographic data file This data file contains the orography and the physiographic data fields related to the soil scheme 16 is also possible to perform a complete ideal case with ideal orography and non trivial surface conditions user can combine the two possibilities with switches included in the namelist NAM REAL and initialize a simulation with a real orography and idealized homogeneous surface fields If PREP_PGD file is specified and if the switches in namelist NAM REAL set to FALSE homogeneous values can be imposed by the
154. binaries problems could occur when you modify a MODD present in the Master library without compiling all the sources which use this MODD The typical example is to add variable in MODD you must in this case compile again all the sources which use this MODD even if the new variable is not used To prevent this type of problem it is prohibited to modify MODD or MODN which is present the Master library MODN is the 7 3 THE PREPSOURCE PROCEDURE 201 generic name for the module defining the Namelists A control is realized by the prepsource procedure and if this rule is transgressed the procedure stops and nothing is compiled Therefore if you want to add a variable to a given module MODD EXAMPLE create a new module MODDB EXAMPLE which only contains the new variable The proximity of the name MODDB_EXAMPLE with the pre existing module will help the Meso NH administrator to merge the modifications coming from different Meso NH users for the next release of the model Another way is to use the variables available in the module MODD BLANK also in MODN BLANK present in all the MesoNH namelist files by adding just the call to it USE MODD_BLANK in your subroutine see the example of modeln f90 in section 7 4 illustrate the locations of the libraries we plot the different libraries and transfers on the figure 7 1 The procedure prepsource is now described in detail the control parameters are list
155. cal file it is used only if the atmospheric file is a GRIB file It can be either an GRIB file obtained from extractarpege e g an file from the Mocage french model or a MESO NH file obtained in a previous simulation for example the file REALI nam which contains the directives for PREP REAL CASE e Output the MESO NH FM file 5 7 1 The prepmodelrc file PREP REAL CASE is run with the procedure prepmodel The input and output host direc tory and login control variables refer to all the input and output files and not only to input or output FM files The other control variables to initialize specifically are e MAINPROG PREP REAL CASE e NAMELISTFILE default e LISTGET default These variables contain the names of the input atmospheric file and Physiographic data FM file respectively 160 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES 5 7 2 The physiographic data file This is a FM file but with fewer elements than a MESO NH file It contains the physiographic 2D fields The geographic and grid data are stored on this file This file is created by the program PREP_PGD It is possible to use a complete MESO NH file since it also contains the physiographic fields If one wants only modify the vertical grid of a MESO NH file without any change on the orography one can specify it both as atmospheric file and physiographic data file It contains e the definition of the projection the horizont
156. cal switch is set to TRUE the radiation computations are made for one mean clear sky and for the whole cloudy columns This is still the way to spare some CPU time by postulating that the clear sky columns do not lead to very different radiative tendencies This hypothesis is only valid in academical cases e NRAD COLNBR Maximal number of air columns called in a single call of the radiation subroutine This is performed in order to save memory because the radiation subroutine allocate for every column of size NKMAX NKMAX working arrays This leads to a quadratic dependency of the memory with the number of vertical levels of the model A way to limit the necessary memory is to split the number of columns passed to the radiation subroutine in several sets of NRAD COLNBR column Finally all the desired columns depending on the preceding parameters will be treated by sequentially calling the radiation subroutine for every set of column e NRAD DIAG number of diagnostic fields related to the radiative scheme stored on every output synchronous files same fields as NRAD 3D in DIAG program p 184 e XFUDG subgrid cloud inhomogeneity factor The cloud overlap assumption is defined in the routine ini radconf f90 The different as sumptions are e NOVLP 5 Random overlap for Clear Sky fraction and Effective Zenithal Angl It is the best choice without subgrid condensation e NOV LP 6 Maximum Random Overlap for Clear
157. called with crontab INPUT files and OUTPUT file 2 1 PRESENTATION 15 DEBUGSCRIPT switches command echos on or off for an easy script debugging ON or OFF this debugging level reports to the procedures and not to the Fortran 90 subroutines compilation PATH the pathname to find the Meso NH procedures must be added to PATH For example PATH PATH mesonh procedures MANPATH the path to find the man files must be added EDITOR this is the ascii file editor often set to vi SIMUL absolute name path directory name for the simulation directory on which the MESONH procedures will work it contains extracted source files namelist files specific files containing personal default values used by the procedures 2 1 3 The input parameters of MESONH script The meaning of the control parameters needed by every main MESONH script is described in the next section We only present here the different ways to set up these parameters e They can be introduced via variables stored in input files named scriptrc script stands for the name of any procedure and rc is the suffix If scriptrc is not provided in the SIMUL directory then a default file in the Meso NH root directory 6MESONH procedures will be searched These files are strongly recommended because they are obligatory in a non interactive session and used to provide own defaults for an interactive session To create your own version of the scr
158. ces actually performed on REMOTE HOST to compile the source files of step 1 An example is presented in the next section where new subroutine is compiled with a modified source of the Meso NH model 204 CHAPTER 7 MODIFY THE FORTRAN SOURCES 7 44 An illustrative example of a source compilation We first give the Fortran source of the new routine in order to illustrate the Meso NH style for the routine cfl its interface encapsulated in a module modi cfl and the subroutine model n present in the Master library FILE cfl f90 MODULE MODI CFL HHHHHHHHHHHHHHH INTERFACE SUBROUTINE CFL PUT PVT PDXHAT PDYHAT PTSTEP HLUOUT TPDTCUR USE MODE TIME REAL DIMENSION INTENT IN PUT PVT 2 components of the horizontal Wind at t REAL DIMENSION INTENT IN PDXHAT Stretching in x direction REAL DIMENSION INTENT IN PDYHAT Stretching in y direction REAL 5 INTENT CIN PTSTEP time step CHARACTER LEN INTENT CIN HLUOUT Name of the output listing TYPE DATE_TIME INTENT OUT TPDTCUR Current date and time END SUBROUTINE CFL END INTERFACE END MODULE MODI_CFL SUBROUTINE CFL PUT PVT PDXHAT PDYHAT PTSTEP HLUOUT TPDTCUR
159. cle de donnees CIT 22378 mots position 698372 a 720749 article de donnees RST 22378 mots position 720750 a 743127 article de donnees RGT 22378 mots position 743128 a 765505 article de donnees DRYMASST d 23 mots position 765506 765528 article de donnees SRCM 22378 mots position 765529 787906 article de donnees SRCT 22378 mots position 787907 810284 article de donnees SIGS 22378 mots position 810285 a 832662 article de donnees RHOREFZ S 68 mots position 832663 a 832730 article de donnees THVREFZ 68 mots position 832731 a 832798 article de donnees EXNTOP n 23 mots position 832799 a 832821 article de donnees INPRR wy 508 mots position 832822 a 833329 article de donnees ACPRR w 508 mots position 833330 a 833837 article de donnees INPRS 508 mots position 833838 a 834345 article de donnees ACPRS n 508 mots position 834346 a 834853 article de donnees INPRG S 508 mots position 834854 a 835361 article de donnees ACPRG WS 508 mots position 835362 a 835869 article de donnees INPRT 508 mots position 835870 a 836377 article de donnees ACPRT ur 508 mots position 836378 a 836885 article de donnees PABST 22458 mots position 836886 a 859343 article de donnees NEB 22458 mots position 859344 a 881801 b giri 81 articles logiques de donnees et 1 trous repertories listes LFILAF Fin du catalogue de l Unite Logique 12 82 Articles logiques en tout
160. ctiva tion spectrum s e XCONC CON aerosol number concentration N e XR MEAN geometric mean radius of the aerosol distribution 7 e XLOGSIG CCN natural logarithm of the geometric standard deviation of the aerosol distribution In c e XFSOLUB_CCN Mean solubility of the aerosols em e XACTEMP_CCN Mean air temperature at which activation will occur e XALPHAC First dispersion parameter of the y distribution law of the cloud droplets v D TO Age eae AcDJ e XNUC Second dispersion parameter ve of the y distribution law of the cloud droplets e XALPHAR First dispersion parameter o of the y distribution law of the rain drops T OrVr 1 ar 60 p Qv Dy e XNUR Second dispersion parameter vp of the y distribution law of the rain drops e LRAIN Enables the rain formation by cloud droplet autoconversion when it is TRUE e LSEDC Cloud droplets are allowed to sediment when it is TRUE e LACTIT Activation by radiative cooling is taken into account when it is TRUE 4 2 9 Namelist NAM option for the mixed phase cloud pa rameterization ICE3 and ICEA Fortran type default value LWARM logical CPRISTINE ICE 4 characters PLAT LSEDIC boolean FALSE It contains the options for the mixed phase cloud parameterizations used by the model They are included in the declarative module MODD PARAM ICE e LWARM When
161. d now we read both 42 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE 04 real K at ground level Nota it is used later in the code to compute if asked a time varying sea surface temperature ry real kg kg at ground level number of level integer height of levell real m if ZFRC or pressure at levell real Pa if PFRC Ufre component at levell real m s component at levell real m s 0 fre at levell real K Ty fre at levell real kg kg Were at levell real m s 00 Ot fre at levell real K s and Or Ot fre at levell real 1 s 00 02 fre at levell real K m 80 09 fre at levell real K m e idem at level2 e idem at levelN If PFRC is the forcing type an additional sounding is given in order to convert the pressure levels into height levels with enough accuracy Data are organized as follows e number of level integer e pressure at levell real Pa 0 at levell real K and r at levell real kg kg This operation is repeated until the previous number of sounding is reached Example of free part of PRE IDEA1 nam ZFRC 1 1983 07 01 O 3 4 THE OUTPUT MESONH FILE 43 0 284 5 008 6 7 0 0 0 281 10 0 00540 0 00000 0 0 0 0 15 7 0 0 0 281 10 0 00540 0 00000 0 0 0 0 1095 7 0 0 0 280 75 0 00540 0 00300 0 0 0 0 1145 7 0 0 0 290 60 0 00190 0 00300 0 0 0 0 3000 7 0 0 0 304 15 0 00190 0 00300 0 0 0 0 9000 7 0 0 0 346 1
162. d example of 6 2 3 the output files will contain the set of variables X000 Y000 Z000 000 RV000 for origin corresponding to the last file CFILES 6 and 2 extra sets Xn Yn Zn THn RVn with n 001 for origin corresponding to the CFILES 4 n 002 corresponding to CFILES 2 Note that extra origins are in chronological order 5 Namelist NAM PARAM CONVECTm options for the convective scheme when convective diagnostics with NCONF_KF see chapter 4 for variables meaning 6 Namelist NAM PARAM RADn options for the radiations when radiation diagnostics with NRAD 3D see chapter 4 for variables meaning 7 Namelists of the externalized surface Diagnostics of the surface parameters are not computed by MESONH itself but by the externalized surface included in it So you are invited to refer to the documentation of the surface e namelist NAM DIAG SURF ATMn e namelist NAM DIAG SURFn e namelist NAM DIAG ISBAn namelist NAM DIAG TEBn namelist NAM DIAG CSEAn namelist NAM DIAG WATERn 6 2 3 Examples of 1 e Namelist file for 2 files DIAG LVAR LS T NCONV_KF 2 6 2 EXECUTION 181 NRAD_3D 1 CRAD_SAT METEOSAT LVAR_MRW T LVAR MRSV T LMOIST V T LMOIST E F LTPZH T LVORT F LMSLP F LGEO T LAGEO T LWIND ZM F LTHW T LCLD_COV T LVAR PR F LTOTAL PR F LMEAN PR F XMEAN PR 1 2 4 NCAPE 1 LRADAR T LTRAJ F amp NAM DIAG BLANK amp NAM DIAG FILE YINIFILE 1
163. del 1 The extraction must be done separately for each date and time for the initial file and each of the coupling file of model 1 4 PREP_REAL_CASE this program is run several times for the initial file and the coupling files of model 1 5 MESONH this step is optional If you do not wish to start all the models at the same time you can decide to run the model 1 before the model 2 starts 6 ZOOM_PGD Since the second PGD file was done for models 2 and 3 you have to zoom it on the domain of model 2 with this program 5 1 OVERVIEW OF THE INITIALIZATION SEQUENCES 143 7 SPAWNING when you want to start the model 2 you must use this program to compute the horizontal interpolations from the model 1 to the model 2 It is used only once for the initialisation of model 2 8 PREP REAL CASE It is used only once to compute the initial file for the model 2 Do not change the vertical grid 9 MESONH here is your complete nested run with model 1 and model 2 10 ZOOM PGD Since the second PGD file was done for the models 2 and 3 you have to zoom it on the domain of model 3 with this program The domain of model 3 has a common zone with the one of model 2 11 SPAWNING when you want to start the model 3 you must use this program to compute the horizontal interpolations from the model 1 and to use the fields of model 2 in the common domain It is used only once for the initialisation of model 3 12 PREP REAL CASE It is used on
164. dient for non symmetric problems with the same preconditioner CREST Conjugate Residual method If the problem to be solved is flat and cartesian then the resolution becomes exact and no iteration is performed 66 CHAPTER 4 PERFORM A MESONH SIMULATION NITR Number of iterations for the iterative pressure solver The value of this parameter depends on the maximum slope of the orography present in the model LITRADJ Logical to adjust the number of iterations for the iterative pressure solver according to the range of the residual divergence XRELAX Relaxation coefficient in the Richardson method CPRESOPT RICHA This value be less than 1 only for very steep orography in general the optimal value is equal to 1 LHORELAX_UVWTH Switch for the horizontal relaxation applied on the outermost verticals of the model for U V W variables TRUE The horizontal relaxation is applied FALSE The horizontal relaxation is not applied LHORELAX RV LHORELAX RC LHORELAX RR LHORELAX RI LHORELAX RS LHORELAX_RG LHORELAX RH LHORELAX TKE LHORELAX SV LHORELAX SVCHEM LHORELAX_SVC2R2 LHORELAX SVCIR3 LHOREAX SVLG LHORELAX SVDST LHORELAX SVAER LHORELAX_SVELEC idem for other variables It is more safety to set all the LHORELAX values than use the default values which can be modified by the desfm file LVE_RELAX Switch for the vertical relaxation applied on the outermost verticals of
165. domain Example of namelist PRE ZOOMI nam amp NAM PGDFILE CPGDFILE PGDFILE_1 neste1 YZOOMNBR 258 amp NAM MESONH DOM 60 NJMAX 50 NXOR 5 NYOR 8 152 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES 5 5 Extraction of ECMWF files extractecmwf extractecmwf prepares job which will perform the extraction of meteorological data at ECMWF using MARS Meteorological Archival and Retrieval System Fields are on Gaus sian grid or on lat lon regular grid if limited area is requested in a standard GRIB file It can be handled with C routines such as in library pbio to make the interface with fortran programs To treat its information one needs the library gribex Remark The extraction is executed on ecgate you need an account on it to submit the job either directly during a session on ecgate or through an ECaccess gateway extractecmwf has control variables are given in the file extractecmwfrc or interactively They are e OUTHOST output file will be put or transfered onto a storage machine ecfs ECMWF local storage system ftp gw transfer protocol through gateway to the Idris storage machine gaya user and password have to be specified in the HOME netrc file of ecgate ectrans ECaccess command via Internet for Meteo France storage machine cougar e OUTLOGIN user on the OUTHOST machine to dispose the desarchived GRIB file e OUTDIR name of directory on the
166. e and the coupling files In this case the sequence uses figures 5 3 and 5 4 1 PREP_PGD 2 SPAWNING this program performs the horizontal interpolations from a MESO NH file into another MESO NH file with a finer resolution and smaller domain 3 PREP REAL CASE 5 1 3 Initialization for a nesting run Two or more MESO NH models will be run interactively The first model the one with the coarser resolution and containing the others needs again one initial file and some coupling files for lateral boundary conditions However the nested models are contained in another MESO NH model This allows to give them their lateral boundary conditions directly interpolated from the model which contains it at all time steps Therefore no coupling file is necessary for those models The initial file must still be computed before the run the interpolation program for all the model fields is not yet included in the model program itself The user can choose the date of the nested model start as he wants it is not necessary the same than the coarser model The only obligation is to start a nested model only if all the coarser models containing it have already started or are starting The initialisation sequence is a merging of the two previous ones initialisation and coupling files of the first model initialisation files from MESO NH model for the nested files However there is major change at the beginning of the sequence all t
167. e as in the previous examples Lon ean eats FULL SCREEN VISUALISATION OF THE CONTROL PARAMETERS TP the user can change them here 1 gt tosupc script 4 EXAMPLE OF THE TOSUPC SCRIPT 227 Copy or creation f90 o I BIBUSER necessary shell scripts ABSMOD exe namelists input FM files if DEBUG cdbx copy namelists f90 input FM files ABSMOD exe this allows interactive debugging on workdir tmpdir execdir copy namelists input FM files ABSMOD exe after execution ABSMOD exe run ouput FM files are stored on Supercalc or workdir otherwise on Archive and on workdir OUTDIR Figure 2 The objects that are copied or created on temporary and work directories during a Meso NH run 228 APPENDIX A AN ILLUSTRATIVE EXAMPLE OF USE OF PROCEDURES 1 gt simulation directory SIMUL users come fischer mnh master 2 gt 05 toto XX NORM 1200 384 1 no 1 2 gt SUPER CALCULATEUR D EXECUTION 2 gt 2 gt passage par xoper NQS vers 2 tora ou toba pour le VPP5000 2 gmor pour le 5000 class getmor 2 gmax pour le VPP5000 class getmax 2 toco pour le VPP5000 class compil 2 pour les super calculateurs non meteo 2 idris pour l idris passage par hebe 2 par defaut tora kami Step 1 gives the possibility to the user to change the control parameters which can however also be initialized directly in file execfilerc In step 2
168. e files 155 5 7 Vertical interpolations PREP REAL CASE 159 Otel Fhe prepraodelrc hle d SR ede reo ib RD Es 159 5 7 2 The physiographic data file 160 Ehe atmospheric hle Mob ee a 160 5 7 4 The chemical file optional 160 5 7 5 1 160 5 8 Horizontal interpolation from a MESO NH SPAWNING 169 5 81 prepmodelrc 169 CONTENTS 5 7 5 8 2 The input SPAWNI nam fle 169 5 9 If you want surface fields only PREP SURFEX 172 5 9 1 172 5 9 2 The physiographic data file 172 5 9 9 Ehe anput file wm i hoe a Oe ER s 173 5 9 4 The file 1 173 5 10 Miscellaneous interactive procedures 175 5 10 1 Content of a MESO NH file fmmore 175 5 10 2 Definition of the vertical grid 175 5 10 3 Conversion from spherical coordinates to conformal coordinates latlon2xy175 5 10 4 Conversion from conformal coordinates to sphericalcoordinates xy2latlon 176 Compute diagnostics after MESO NH simulation 177
169. e input 4 21 Namelist NAM BLANK available variables 4 2 2 Namelist NAM CONF global configuration 4 23 Namelist NAM DUST 6e e epos a OR qos Se ede x SOR 4 20 4 Namelist NAM DYN global parameters for the dynamics 4 25 Namelist NAM FMOUT output 4 2 6 Namelist NAM FRC forcing control 4 2 7 Namelist NAM NESTING grid nesting configuration 4 2 8 Namelist NAM PARAM C2R2 control variable of the 2 moment warm microphysical schemes C2R2 and 4 2 9 Namelist NAM PARAMLICE option for the mixed phase cloud parame terization 1 and ICE4 4 2 10 Namelist NAM TURB CLOUD mixing length for clouds 4 2 11 Namelist NAM ADVn scalar advection schemes of model n 4 2 12 Namelist NAM CONFn configration 1 4 2 13 Namelist NAM DYNn parameters for the dynamics of model n 4 2 14 Namelist NAM LBCn boundary conditions of model n 4 2 15 Namelist LUNITn file 4 2 16 Namelist NAM NUDGINGn nudging of model n 4 2 17 Namelist NAM PARAMn parameterizations names of model n 34 51 51 CONTENTS 4 3 3 4 2 18 Namelist NAM PARAM CONVECTn op
170. e master RCS library are prohibitted Let us now consider the case where ENVIRONMENT SILENTINTERACTIVE Then mesonh is not asking anything to the user that s why it is silent but it still prints information on the screen so it keeps some interactive shape In fact prepsource is now reading input from a file that must be called inprepsource and be present in the catalog SIMUL This file contains the following information see the prepsource man page filel VERSION dir2 file VERSION dir3 file3 VERSION compiler options Each information separated by commas is stored on one line Thus most part of inprepsource consists of pairs of lines containing first directory name and second a file and version name This is in fact the input that prepsource asked for in steps 4 and 5 of the previous example The last two lines contain a separator string gt followed by the compiler options which were asked for in step 12 above Such a file inprepsource can of course be directly written by the user However prepsource itself generates during each session a file called inprepsource new that may be used in a future session after being copied to inprepsource When ENVIRONMENT BATCH prepsource is running without any contact to the user In this case output is directed onto a temporary file and no input is read at all mesonh simply takes all source files that are stored SSIMUL and prepares them for remote compilation
171. e right executable binary ABSSMAINPROG exe in the work directory on the remote host where the model executions are performed The chapter 7 will explain these steps different localizations and information transfers are illustrated in figure 2 2 vertical line separates what happens in the local host from the actions performed in the remote host The first part of an experiment consists in the preparation of a file named outscript which is composed of a list a UNIX orders and a copy of the pertinent namelists This file is transfered in the remote host where it is executed this job is self governing and is able to extract the required file either from the remote host or from a storage machine linked to it The different outputs listings FM synchronic files FM diachronic files are disposed at the rigth destinations during still the same job included in the outscript file local host remote host storage machine Before the extensive presentation of the different steps of the prepmodel and tosupc 2 1 PRESENTATION 13 Qd J L eee due uud prep one MesoNH procedure script prep or extract or tosupc 1 i outprep output file from prep or extract tosupc to execute outprep file output file automatically executed Figure 2 1 MESONH procedures procedures we briefly describe the different ways to get informations either by environment
172. ea surf vp 221 6 15 2 where z is the orography and T7 is the mean virtual temperature between the ground level and the sea level the latter is extrapolated from the first with a climatological gradient of 6 5K km 190 CHAPTER 6 COMPUTE DIAGNOSTICS AFTER A MESO NH SIMULATION Thickness of water species THVW THCW THRW THIC THSN THGR THHA The thickness of a water specy x with 2 v c r i s g or h is computed as k kp Pavel 6 16 Height of explicit cloud top HEC For every columns scanned from the model top to the bottom the height of explicit cloud top is the height where the cloud mixing ratio exceeds the value of 0 1g kg If a mixed microphysical scheme is activated during the simulation the ice mixing ratio r is also taken into account with the same threshold and the height is the higher between the one of the top determined with re and the top determined with r Height and temperature of maximum cloud top HC TC If a convection scheme is activated during the simulation and if you ask for convective diagnostics NCONV_KF gt 0 the top of convective cloud computed by the convection scheme is compared to the previous one of explicit cloud in every columns The height and the temperature of the higher top are deduced For clear sky columns the height is 0 and the temperature is the one of the ground Visibility VISI visibility function of the liquid water conten
173. eal in Kelvin the liquid potential temperature in case KIND ZUVTHLMR real in Kelvin The moist variable is the dew point temperature in case KIND STANDARD real in Kelvin the vapor mixing ratio in cases KIND PUVTHVMR or ZUVTHDMR or ZUVTHVMR or PUVTHDMNR real in Kg Kg the total water mixing ratio in cases KIND ZUVTHLMR real in Kg Kg the relative humidity in cases KIND ZUVTHVHU or PUVTHDHU or PUVTHVHU real in percents Additional cloud variables For the moment this configuration works only for KIND PUVTHDMR or ZUVTHDMR and LID TRUE It is planned to compute radiation diagnostics with the DIAG program see chapter 6 cloud mixing ratio if LUSERC T or LUSERI T real in Kg Kg ice mixing ratio if LUSERI T real in Kg Kg 38 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE You should take care that the levels are dense enough so that the Laplace relation which gives the thickness between successive levels can be applied The radiosounding informations are written in the file in the following order e YEAR integer exemple 1994 MONTH integer exemple 4 DAY integer exemple 22 TIME real in seconds exemple 36000 for 10 h e KIND of data used for the radiosounding string of 8 charcaters Nine kind of data are possible STANDARD PUVTHVMR PUVTHVHU ZUVTHVMR ZUVTHVHU PUVTHDMR PUVTHDHU ZUVTH
174. ecifically in this file are e MAINPROG DIAG e NAMELISTFILE default e LISTGET default this variable contains the names of the input FM files up to 24 e OUTHOST name workstation for example this allows future use of diaprog on your workstation if you chose to convert into a diachronic file 6 2 2 The namelist file DIAGI nam The DIAG1 nam namelist file contains the diagnostics required by the user the name of the input FM files the suffix of the output diachronic files and output file type The user can reset options for the convective and radiation scheme with NAM PARAM CONVECTn and NAM PARAM RADn namelist see chapter 4 The order of namelists is free and unset namelists can be ommited 1 Namelist NAM DIAG controls diagnostic variables CISO character 1 6 PREVTK LVAR RS logical TRUE all other logicals FALSE 257 all other character strings all integers 1 LDIAG array of logicals 100 FALSE XDIAG array of reals 100 XUNDEF 6 2 EXECUTION 179 see section 6 3 for meaning Fortran type default value LAIRCRAFT_BALLOON logical FALSE NTIME_AIRCRAFT_BALLOON integer NUNDEF XSTEP_AIRCRAFT_BALLOON real XUNDEF XLAT_BALLOON array real 9 XUNDEF XLON BALLOON array real 9 XUNDEF XALT_BALLOON array real 9 XUNDEF e LAIRCRAFT_BALLOON flag to compute aircraft and balloon trajectories with sta tionnary fields Trajectories will be writen in diachronic file YINJFILEBAL e NTIME_AIRCRAFT_BALL
175. ed below e Parameters concerning the LOCAL HOST MKFNAME name for the model s makefile default MESONH procedures make_mnh practically always used e Parameters concerning the REMOTE HOST BIBMASTER file name for master binary library on the REMOTE HOST BIBBUGFIX file name for bugfix binary library on REMOTE HOST BIBUSER file name for the user s binary library on the REMOTE HOST starting at or 0 if none e Other control parameters DEBUG compilation with debugging option or not debug or run on the RE HOST NSOURCE number of copies for multi tasked code 1 to 8 Default values can be provided by a personal version of the prepsourcerc file which can be taken in the directory mesonh procedures There are eight possible configurations depending on BIBMASTER BIBBUGFIX and BIBUSER e BIBMASTER something a BIBBUGFIX 0 and BIBUSER 0 In this case all object code that is not generated explicitly during the prepsource step will be extracted from the master binary library T his configuration allows to work only with the master library and no user code e BIBMASTER something a BIBBUGFIX bugfix name a and BIBUSER 0 In this case the object code will be extracted from the bugfix binary library if it is present in the master binary library otherwise 202 CHAPTER 7 MODIFY THE FORTRAN SOURCES L
176. ee and unset namelists can be ommited 1 Namelist NAM FILE NAMES contains file names dest vals HATMFILE character LEN 28 HATMFILETYPE character LEN 6 GRIBEX HPGDFILE character LEN 28 CINIFILE character LEN 28 INIFILE e HATMFILE name of the atmospheric file up to 28 characters e HATMFILETYPE type of the atmospheric file GRIBEX MESONH e HPGDFILE name of the Physiographic Data File up to 28 characters e CINIFILE name of the MESO NH output FM file used as initial or coupling file in a MESO NH simulation 2 externalized surface namelists for PREP_SURFEX surface initial fields are produced by externalized surface facilities So you are invited to refer to the documentation of the surface For PREP_SURFEX you must fill the following namelists e NAM_PREP_SURF_ATM e NAM PREP SEAFLUX if you chose to use the SEAFLX scheme e NAM PREP WATFLUX if you chose to use the WATFLX scheme e NAM PREP TEB if you chose to use the TEB urban scheme e NAM PREP ISBA if you chose to use the ISBA scheme 174 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES CAUTION a Note that all namelists can be void but only if the initial file name for HATMFILE you provide in namelist FILE NAMES contains externalized surface fields b If the file HATMFILE does not contains externalized surface fields you must fill at least namelist NAM PREP SURF ATM
177. el to compute diagnostic fields This Chapter will therefore present how to form a job built with one or more elementary steps It should be noted that a very modular structure has been adopted for the procedures and thus a job performing more than one elementary step is only the assembling of the different scripts performing each elementary step script is UNIX procedure which e collects informations from different sources environment variables the content of one or more files containing Fortran subroutines or lists of variables e puts together all theses informations in a single file named outexample for the script example e executes the job included in this file elementary scripts fig 2 1 can perform the following actions e modify your binary library by compiling a subroutine which is added or modified by comparison with the master library including the Bugfix library or not step prepsource the presentation of this step is described in chapter 7 e perform either the preparation of an initial MESONH file either the time dependent sim ulation or any main program present in the master library step prepmodel e execute your complete job step tosupc 11 12 CHAPTER 2 THE MESONH PROCEDURES EXCEPT PREPSOURCE e ask for operational datas from ECMWF step extractecmwf or from Meteo France step extractarpege This user s guide is not the only way to get informations on the procedures the
178. elist description 27 namelist description 28 namelist description 159 namelist description 64 namelist description 28 namelist description 91 namelist description 28 65 namelist description 146 namelist description 57 namelist description 65 INDEX INDEX namelist description 57 NAM PARAM CONVECTn namelist description 67 NAM PARAM ICE namelist description 59 NAM PARAM RADn namelist description 68 NAM PARAMn namelist description 66 NAM PDF namelist description 95 NAM PERT PRE namelist description 29 NAM PGDFILE NAM REAL CONF namelist description 157 NAM REAL PGD namelist description 30 NAM SERIES namelist description 79 NAM SERIESn namelist description 80 NAM SLEVE namelist description 30 NAM_TURBn namelist description 71 NAM VER GRID NAM VPROF PRE namelist description 32 NASSERC NASSERG NASSERH NASSERI NASSERR NASSERS NASSERV NASSESV NASSETH namelist description 141 146 namelist description 30 157 in namelist NAM BU RRC 87 in namelist NAM BU RRG 90 in namelist NAM_BU_RRH 91 in namelist NAM BU RRI 89 in namelist NAM_BU_RRR 88 in namelist NAM BU RRS 90 in namelist NAM BU RRV 86 in namelist NAM BU RSV 91 in namelist NASSETKE in namelist NASSEU in namelist NASSEV in namelist NASSEW in namelist NAUTORC in namelist NAUTORR in namelist NAUTSRI in namelist NAUTSRS in namelist NBERFIRC in namelist NBERFIRI in na
179. emperatures water and snow contents PGD fields stands for cover fractions orographic vegetation and soil physiographic fields and prognostic fields Figure 5 3 schematic view of the interactions between the different files during the initialization sequence of a real case simulation from a MESO NH file 140 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES prepmodelrc PRE_PGD1 nam Creation of PhysioGrafic Input FM file Data file coarse resolution prepmodel PREP_PGD PhysioGrafic Data file fine res Horizontal prepmodelrc interpolations of a MESO NH file nam prepmodel SPAWNING small domain Spawned FM file coarse res fine grid Vertical PRE_REAL1 nam interpolations subdomain prepmodel pai REAL CASE MESO NH file fine resolution Figure 5 4 initialization sequence of a real case simulation from a MESO NH file 5 1 OVERVIEW OF THE INITIALIZATION SEQUENCES 141 e The following initialisation and gridnesting sequence is shown here for three models model 2 included in model 1 and model 3 included in model 2 figure 5 5 qi 10 11 PREP PGD this program is run as many time as the number of models one physiographic data file for the model 1 definition of projection resolution domain one physiographic data file for the model 2 same projection definition of reso lution domain one physiographic
180. en 2 700 YDADINIFILE character len 28 YDADSPAFILE character len 28 YSONFILE character len 28 CINIFILE name of the initial FM file 1 father domain which will be used to spawn model 2 YDOMAIN name of the file which defines the domain for model 2 If domain file is provided for YDOMAIN then all the information of namelist NAM GRID2 SPA will be ignored YSPAFILE optional name of the spawned FM file 2 output file If the user does not specify this name or if YSPAFILE CINIFILE the code builds the spawned FM file name as YSPAFILE CINIFILE spaYSPANBR or YSPAFILE CINIFILE sprYSPANBR if YSONFILE is provided YSPANBR NumBeR which will be added to CINIFILE to generate the FM file name of the SPAwned file string of 2 characters YDADINIFILE if GBAL ONLY TRUE name of the dad of CINIFILE 5 8 HORIZONTAL INTERPOLATION FROM A MESO NH FILE SPAWNING 171 e YDADSPAFILE if GBAL_ONLY TRUE name of the dad of YSPAFILE Pro gram will check that YDADINIFILE and YDADSPAFILE have the same character istics before replacing the dad name of YSPAFILE by YDADSPAFILE instead of YDADINIFILE YDADSPAFILE must exist before running the spawning job e YSONFILE optional name of a spawned FM file input file It must have the same resolution as the spawned FM file 2 output file The fields of YSONFILE will be used at points included in the domain defined by YDOMAIN or NAM GRID2 SPA instead of in
181. ent in the file source f90 in the new version of the library nh takeout source f90 VERS1 takes out of the library the VERS1 version of source f90 and put it in the file source f90 nh unextract source f90 VERS1 undoes the previous extraction no new 22 nh unregister source f90 VERS1 undoes the previous registration the ee nh show VERS1 shows the version identification numbers for all the files present in the library corresponding to the accronym VERS1 nh story source f90 VERS1 lists the historic of the source file starting from the RCS version that corresponds to the acronym VERS1 nh target source f90 VERS2 creates the acronym VERS2 the last version identification number is the one present in the file v Table 7 2 Full syntax of the Meso NH macros and result of their actions Supplementary examples are given in the appendices of this book Global mode behaviour the Meso NH macros may also be used on a whole directory In this case the user must give a directory name instead of a file name in the macro arguments Each macro will then perform its action on all source files in the directory For example nh extract users come fischer masteri VERS1 will extract all the source files associated to VERS1 and present in master1 Furthermore the source code is collected into a global source file named master1 f90 Global mode behaviour is accepted by 200 CHAPTER 7 MODIFY THE FORTRAN SOURCES nh extract
182. er XBOXWIND half width of the box inside which the dynamical center is searched from the guessed position km XRADGUESS guess of the radius of the domain in which the cyclone will be filtered km NPHIL number of azimuthal directions used for the cylindrical coordinates 5 7 VERTICAL INTERPOLATIONS PREP_REAL_CASE 165 NDIAG FILT allow storage of several components calculated from total fields Be careful the components are on the Grib vertical grid in diaprog plot them only on _K_ levels Then to visualize all the Grib vertical levels the number of MesoNH vertical levels must be equal or greater than the number of levels in the input Grib file 0 total unfiltered fields UT15 15 for wind components PRESTOT for absolute temperature and surface pressure environmental filtered fields total field minus hurricane disturbance com ponent UT16 VT16 TEMPENV PRESENV 0 1 basic fields low pass component isolated by the Barnes filter UT17 VT17 TEMPBAS PRESBAS 0 1 2 total disturbance tangential wind component XVTDIS LBOGUSSING to switch on the addition of the bogus vortex logical XLATBOG latitude of the bogussed position of the analytical cyclone center XLONBOG longitude of the bogussed position of the analytical cyclone center XVTMAXSURF maximum tangential wind near the surface or about 500 m altitude m s XRADWINDSURF radius of maximum wind near the surface or about 500 m alt
183. er NADVRV total advection integer NADVXRV advection along x integer NADVYRV advection along y integer NADVZRV advection along z integer NFRCRV forcing integer NNUDRV nudging integer NDIFRV numerical diffusion integer NRELRV relaxation integer NDCONVRYV deep convection integer NHTURBRV hori turb diffusion integer NVTURBRV vert turb diffusion integer NREVARV rain evaporation integer NCONDRV vapor condensation integer or cloud water evaporation NHENURV heterogenous nucleation integer NDEPSRV deposition on snow integer NDEPGRV deposition on graupel integer NCDEPIRV cond deposition on integer nj p 9 cOcococccoccocococococococccccoo 90 4 2 36 LBU RRC NASSERC NNESTRC NADVRC NADVXRC NADVYRC NADVZRC NFRCRC NDIFRC NRELRC NACCRRC NAUTORC NSEDIRC NCONDRC NHONRC NRIMRC NWETGRC NDRYGRC NIMLTRC NBERFIRC NCDEPIRC NHENURC NSEDIRC NWETHRC NDCONVRC NHTURBRC NVTURBRC CHAPTER 4 budget switch time filter Asselin nesting total advection advection along x advection along y advection along z forcing numerical diffusion relaxation deep convection hori turb diffusion vert turb diffusion accretion autoconversion into rain sedimentation of cloud vapor condensation or cloud water evaporation homogeneous nucleation ICE3 or ICEA riming of cloud water ICE3 or ICEA wet growth of graupel ICE3 or ICE4 dry growth of graupel or ICE4
184. esults BSPLITTING domain is decomposed in Box along X and Y XSPLITTING the X direction is splitted in stripes along Y YSPLITTING the Y direction is splitted in stripes along X e LLG Switch to use LaGrangian variables e LINIT_LG Switch to reinitialize LaGrangian variables e CINIT LG When reinitialize LaGrangian variables FMOUT each time an output file is written 42 THE INPUT EXSEG N NAM FILE 55 other string only when starting a new segment CCONF RESTA e LNOMIXLG Switch to not use turbulence for LG variables e CEXP Experiment name this is the name of the set of run you have performed or you want to perform on the same physical subject Please do not leave any blank character in this name e CSEG Name of segment this is the name of the future run you want to perform Please do not leave any blank character in this name From these two last informations we built the names of the different MESONH output files CEXP n CSEG nbr where n represents the number of the model which generates this output and nbr is the number of the outfile For instance if HY DRO and CSEG INIT1 and we use only one model no gridnesting the different output will be called HY DRO A INIT1 001 HY DRO 1 INIT1 002 4 2 3 Namelist NAM DUST This namelist is use to activate explicit aerosols dusts It is not necessary to use chemistry to activate dusts but it is recommended to activa
185. evel 9 Second free formatted part related to chemical species This part is only used if you have previously run extractarpege with MOCAGE outputs This part has to be written at the end of the namelist file In this case the list of the MesoNH chemical species and their corresponding grib code in the Grib file is specified as follows MOC2MESONH transfer mocage RACM variables default values 2 NUMBER OF OPTIONAL GRIB VARIABLES A4 1X I5 03 180 N02 183 If you only indicate MOC2MESONH the list of default values corresponding of the default ones for extractarpege is read in MESONH procedures PREMOC1 nam and add to your own PRE REALI nam Examples of namelist file PRE REAL1 nam e Gribex file levels being calculated and chemical species FILE NAMES HATMFILE ALT90101500 HATMFILETYPE GRIBEX HPGDFILE PGDFILE 10km CINIFILE example1 amp NAM REAL CONF CEQNSYS LHE NVERB 7 VER GRID NKMAX 60 YZGRID_TYPE FUNCTN ZDZGRD 50 ZDZTOP 500 ZZMAX STRGRD 3000 ZSTRGRD 2 ZSTRTOP 6 amp NAM BLANK MOC2MESONH transfer mocage RACM variables 2 NUMBER OF OPTIONAL GRIB VARIABLES A4 1X I5 03 180 N02 183 168 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES N B the mocage part is written at the end of namelist file MESONH file and levels given manually FILE NAMES HATMFILE 0103 1 01 0017 HATMFILETYPE MESONH
186. evels x t e LMASKLANDSEA switch to separate sea and land points in temporal series t and 2 0 e LWMINMAX switch to compute minimum maximum of vertical velocity W in tem poral serie t See also the namelist NAM SERIESn Some examples of temporal series are available which treat pronostic fields averaged or not vertically If one wants other fields for example diagnostic fields such as relative humidity follows extrema the following Fortran routines must be modified e ini series f90 for initialization of size and name of diachronic records 42 THE INPUT EXSEGS N NAM FILE 83 e seriesn f90 to store and eventually vertically average values during the run e write seriesn f90 to horizontally average and write series in diachronic file 4 2 28 Namelist NAM SERIESn temporal series in diagnoctic file of model n Fortran type default value NIBOXL NIBOXH NJBOXL NJBOXH NKCLS NKCLA NKLOW NKMID NKUP NBJSLICE NJSLICEL NJSLICEH NFREQSERIES integer integer integer integer integer integer integer integer integer integer array 20 integer array 20 integer integer 2 b2 b2 C2 b2 C2 N c N 20 3 XSEGLEN 100 60 e NIBOXL NIBOXH NJBOXL NJBOXH lower and upper indexes along x and y axes of the horizontal box used to average the series t and z t e NKCLS NKCLA K level respectively in the CLS and CLA x t series of U Rv Rr at KCLS
187. execution the output file outprepideal by tosupc outprepideal In this case the final file outprepideal will be send in the batch queues of the remote host and the execution listing report will come back in the local host the FM file generated during 3 5 AN ILLUSTRATIVE EXAMPLE OF PREP IDEAL CASE 49 this step will be disposed in the directory mentioned in prepmodelrc OUTDIR If at contrary you do not submit the preparative job by tosupc you can login the remote host get this file from the localhost by ftp and execute it in an interactive session on the remote host in tmpdir directory by the following command outprepideal In this case the only difference is that the report is directly printed at the screen Please note that the number of points is very weak in this example and therefore you do not have prob lems with the memory allocation but for more points you can overshoot the maximum memory available in an interactive session and in this situation you must dispose your outprepideal file in the batch queues of the remote host to obtain more available memory 50 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE Chapter 4 Perform a MESONH simulation 4 1 What does go in and out The MESONH simulation model used previously prepared MESONH files to initialize the prog nostic variables of the model and all the other informations necessary to perform the simulation like the computational grid fo
188. file is used to store all the data necessary to run any step of numerical experiment Three different files are taken into account in the Meso NH project e the synchronous file contains all the values of all the fields allowing a restart of the model and of some diagnostic fields desired by the Meso NH user these informations are obtained at the same instant during the simulation thus they are synchronous e the diachronic file contains time series of informations desired by the Meso NH user They are obtained during more than one time step of the model It is the format in which your file must be in you want to plot it with the graphics software diaprog you can convert a synchronous file into a diachronic one with conv2dia e the physiographic file contains external informations like orography vegetation classes chemical emissions data sets etc 8 2 1 The synchronous file This type of file contains only informations corresponding to the same instant of the simulation it remains open during a whole time step of the simulation and the writing orders can be given from any routine of the model 8 2 THE MESO NH FILES 215 The descriptive part This part is the list of all the namelists of the EXSEG n nam file Thus a complete description of this part is given with the EXSEG n nam description in chapter 4 If the file has been generated during a segment of the model integration the des part contains the different namelists
189. filter Asselin nesting advection along x advection along y advection along z forcing nudging curvature terms Coriolis term gravity term numerical diffusion relaxation hori turb diffusion vert turb diffusion pressure term logical integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer Fortran type default value nj gt Ez Fortran type default value nj gt 2 U 9 oooocjcooocdcuocrooooqooo 87 88 CHAPTER 4 PERFORM A MESONH SIMULATION 4 2 33 Namelist NAM BU RTH budget for Fortran type default value LBU_RTH NASSETH NNESTTH NADVTH NADVXTH NADVYTH NADVZTH NFRCTH NNUDTH NPREFTH NDIFTH NRELTH NRADTH NDCONVTH NHTURBTH NVTURBTH NDISSHTH NNEGATH NREVATH NCONDTH NHENUTH NHONTH NSFRTH NDEPSTH NDEPGTH NRIMTH NACCTH NCFRZTH NWETGTH NDRYGTH NGMLTTH NIMLTTH NBERFITH NCDEPITH NWETHTH NHMLTTH budget switch time filter Asselin nesting total advection advection along x advection along y advection along z forcing nudging ref pressure term numerical diffusion relaxation radiation convection hori turb diffusion vert turb diffusion dissipation negative rain evaporation vapor condensation or cloud water evaporation heterog nucleation homogen nucleation spontaneous freezing deposition of snow depos
190. h zs 9 m mmr 2 gt 2 I 1 2NIZSXXDELTAXY g NJZS XDELTAX XAX XAY in the three dimensional case h Zs 2 Max 5 1 NIZS XDELTAX in the two dimensional case e XAY Widths in meters a along y for orography in case CZS BELL e NIZS Localization in x direction of the mountain center in the case CZS BELL 2 NIZS XDELT AX It refers to a vertical velocity point at the ground NIZS NJZS e NJZS Localization in y direction of the mountain center in the case CZS BELL ys NJZS XDELT AY default is 300 for mountain and 0 for flat orography 28 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE 3 2 10 Namelist NAM GRn PRE soil scheme choice Fortran type default value CSURF NONE e CSURF ground selector NONE no soil scheme will be activated during the future MesoNH simulation we therefore do not need any soil parameters All the namelists of the externalized surface will be ignored EXTE the externalized surface is used allowing for its amazing large choice of surface schemes 3 2 11 namelist NAM LBCn lateral boundary conditions Fortran type default value CLBCX array 2 characters 27 CYCL CLBCY array 2 characters 27 CYCL e CLBCX represent the type of lateral boundary condition at the left and right boundaries along x CLBCX 1 and CLBCX 2 respectively Possible values are
191. h to apply the horizontal diffusion according to Zangl 2002 adapted to mountainous topography No amplitude is applied for this type of diffusion XT4DIFF in NAM DYNn is inactive TRUE This horizontal diffusion is applied FALSE This horizontal diffusion is not applied XALKTOP Maximum value of the Rayleigh damping in s Jat the top of the up per absorbing layer The shape of the absorbing layer is a sin of 2 see the scientific documentation for more details XALZBOT Height in meters in the physical space of the upper absorbing layer base 42 THE INPUT EXSEGS N NAM FILE 57 4 2 5 Namelist FMOUT output instants Fortran type default value XFMOUT array real 8 192 999 e XFMOUT mi is an array of increments in seconds from the beginning of the segment to the instant where the i th fields output on FM files is realized by model m 4 2 6 Namelist NAM forcing control Application of a specific forcing is enabled by a dedicated switch When a Newtonian relaxation is requested the damping time XRELAX_TIME_FRC and the height fixed or physically based above which the forcing is applied XRELAX HEIGHT FRC and CRELAX HEIGHT TYPE must be set Fortran type default valuc LGEOST_UV_FRC logical LGEOST_TH_FRC logical LTEND_THRV_FRC logical LVERT_MOTION_FRC logical LRELAX_THRV_FRC logical LRELAX_UV_FRC logical XRELAX_TIME_FRC real XRELAX_HEIGHT_FRC real CRELAX HEIGHT TYPE
192. he PGD files for all mod els must be computed before PREP REAL CASE program and as a consequence before the first model run Theses PGD files are then checked and conformity between them for gridnesting is imposed the orography of one model in every grid mesh is set equal to the average of any of its nested model orography on the same area 5 1 OVERVIEW OF THE INITIALIZATION SEQUENCES 139 791 74 gt gt horizontal interpolation 7 vertical interpolation u V W 6 p Ts large zs PGD fields MESONH file large domain coarse grid coarse resolution large scale vertical grid data files ecoclimap orography clay sand files any domain allowed surface physiographic fields small domain H definition Y u V W 6 p Ts arge zs PGD fields small domain gt definition MESONH MESONH file 3 small domain Physiographic data file fine grid coarse resolution small domain DOS large scale vertical grid fine grid fine resolution 2 wee E TR QT ELS 2 t 2 MN 2 UU u V W 0 T p fine scale zs fine PGD fields Ts 42 D MESONH initialization or coupling file small domain fine grid fine resolution new vertical grid large zs horizontally interpolated orography large scale orography Ts temperature of surface stands for all other surface variables t
193. iac file conditions uu DW FRAC zi __ downdraft action ws ww T lt 9 gt wm gt wom oa woy iG e uw DWKE 2 Pett eu asw Joy noa DW et re PSS 1 zin zia 107 108 CHAPTER 4 PERFORM A MESONH SIMULATION 4 3 9 LES averaged surface fields field notation in dimen general comments diac file sion conditions lt Wh surf gt nn i a 34 surface sensible flux lt gt LU tent fux eee m gt lt Ug surf gt lt h gt convective velocity if positive surface buoyancy flux RE ma 3 ondary eyer height E bk Tenth TK Edz INT TKE t vertical integrated TKE ct base height CF x total loud cover Toc tr Iwe X r Cond water path VARUWP wm n variance 717 Rat water path INPRR INSEPREC 3 Is prec rate _PRATN PREC INPRR over rainy grids ACPRR e n Aem precip rate Arma of cloud fraction maximum 4 3 LES DIAGNOSTICS 4 3 10 LES 2 points correlations field notation in the dim general comments diac file
194. ical component of Absolute Vorticity s LVORT UM1 VM1 WM1 3D relative vorticity components s 55 POVOM 3D Potential Vorticity PVU CISO0 EV PREVTK POVOV 3D Virtual Potential Vorticity PVU LMOIST V POVOE 3D Equivalent Potential Vorticity PVU LMOIST E MEAN POVO 2D Mean Potential Vorticity PVU LMEAN POVO MEAN POVOV 2D Mean Virtual Potential Vorticity PVU LMOIST V MEAN POVOE 2D Mean Equivalent Potential Vorticity PVU LMOISTE averaged between two isobaric levels in Pa XMEAN_POVO 1 2 UM88 VM88 WM88 3D Geostrophic wind components m s LGEO UM89 VM89 WM89 3D Ageostrophic wind components m s LAGEO MSLP 2D Mean Sea Level Pressure hPa LMSLP THVW THCW THRW 2D Thickness of Vapor Cloud and Rain Water mm LTHW THIC THSN THGR THHA 2D lt gt of ICe SNow GRaupel mm HEC 2D Height of Explicit Cloud top km LCLD COV HC TC 2D maximum Cloud top Height km Temperature C CLDFR VISI HOR 3D Cloud Fraction 2 Visibility m ACPRR INPRR 2D ACcumulated INstantaneous Precipitation Rates LVAR PR explicit Rain mm mm h INPRR3D 3D INstantaneous 3D Rain Precipitation flux explicit Rain m s EVAP3D 3D INstantaneous 3D Rain Evaporation flux x explicit Rain kg kg s ACPRC INPRC 2D ACcumulated INstantaneous Precipitation Rates 13 explicit Cloud mm mm h ACPRS INPRS 2D
195. id flux or co variance 1D lt gt horizontal mean value of resolved flux or co variance 1D 4 3 2 What is available The computed fields have usually at least two dimensions z and t that is they are temporal evolutions of vertical profiles They are always written in the diachronic file Each field have its own group name say When time averaging is asked for the fields are temporally averaged and so lose their temporal dimension and are written under the name A NAME When normalization is asked for this one is made individually on each vertical profile for all times They are written under the name E When both normalization and time averaging are asked for normalization is made first and then time averaging The resulting vertical profiles are written under the name H 4 3 3 LES averaged fields LLES MEAN TRUE 4 3 LES DIAGNOSTICS lt u gt lt v gt lt p gt lt p gt lt 0 gt 0i lt 0 gt lt gt Te lt gt Ti 572 lt Tg gt Th gt lt Sy gt lt VU 72 gt V general notation in the 1 1 conditions dim diachronic file N 2 5 N 2 6 N 2 5 N zt N zt Y 2 6 Y 27 5 R MEANERC zi MEANERR zi PR MEAN SV 101 comments MEAN WIND __ diferent from Vc 3273 v lt lt u gt gt zip
196. ile management with their names actions and RCS associated command Of course since most users are not allowed to modify the Master library sources some macros cannot be used with the Master and Bugfix library but only with their private library see last column equiv ROS command library creates the fly UL nh extract extracts a source in order nh register ci tag modifies the source inside nh takeout co extracts a source without intent ER undoes a previous nh extract undoes a previous nh register nh story lists the historic for a given version name creates a mnh file for a preexisting s file Table 7 1 The Meso NH source file management tools name RCS associated command brief description and library upon which it can be applied UL for User s library and ML for Master and Bugfix libraries lists the mnh files for a given version name full syntax of these macros is given in table 7 2 7 2 THE MESO NH SOURCE FILE MANAGEMENT 199 full syntax interpretation nh create source f90 VERS1 creates source f90 v where the version VERS1 1 1 is equal to source f90 nh extract source f90 VERS1 extracts the last version 1 15 say from the library to obtain a new version 1 16 and put it in the file source f90 nh extract source f90 VERS1 1 10 extracts the version 1 10 from the library to obtain a new version 1 10 1 1 because 1 11 already exists nh register source f90 VERS1 registers the modifications pres
197. ion mask MAX take NCH VEC LENGTH points take all grid points HOR take horizontal layers VER take vertical columns NCH_VEC_LENGTH number of points for MAX option TSID TSTEP time between two call to write tsld e CCH TSID COMMENT comment for write tsld e CCH TSID FILENAME filename for write ts1d files 4 2 23 Namelist NAM CH ORILAM This namelist is to activate ORILAM chemical aerosols lognormal distribution for Aitken and accumulation mode This parameterization include coagulation intra and inter modal nucle ation sedimentation condensation adsorption of gas phase This parameterization need to be run together with gas chemical phase namelist NAM CH MNHOn For correct representation it is recommended to have in the chemical scheme severals compounds as nitric acid H2S04 or SULF sulphates NH3 ammonium and CO carbon monoxyde 42 THE INPUT EXSEGS N NAM FILE 79 Fortran type default value LORILAM logical LVARSIGI logical LVARSIGJ logical LSEDIMAERO logical LTHERMIJ logical XINIRADIUSI real XINIRADIUSJ real CRGUNIT character len 4 XINISIGI real XINISIGJ real XNOIMIN real XNOJMIN real XCOEFRADIMAX real XCOEFRADJMAX real XCOEFRADIMIN real XCOEFRADJMIN real CMINERAL character len 5 CORGANIC character len 5 CNUCLEATION character len 80 e LORILAM switch to activate chemical aerosol possible if LUSECHE
198. ion near the top of the model 32 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE ZZMAX STRGRD Altitude separating the two constant stretching layers e ZSTRGRD Constant imposed stretching in 96 in the lower layer below ZZMAX_STRGRD e ZSTRTOP Constant imposed stretching in 96 in the upper layer above ZZMAX STRGRD e LSLEVE flag for Sleve vertical coordinate XLENI decay scale for smooth topography in meters e XLEN2 decay scale for smale scale topography deviation in meters 3 2 17 Namelist NAM VPROF variables for CIDEAL CSTN or RSOU Fortran type default value LGEOSBAL logical CFUNU 3 characters CFUNV 3 characters CTYPELOC 6 characters XLATLOC real XLONLOC real XXHATLOC real XYHATLOC real NILOC integer NJLOC integer TRUE ZZZ ZZZ IJGRID 45 0 20000 4 4 e LGEOSBAL Switch to fulfill the geostrophic balance or not TRUE the geostrophic balance is satisfied by the initial fields FALSE the geostrophic balance is not satisfied by the initial fields e CFUNU String of 3 characters describing the type of function which gives the x com ponent of the wind Possible configurations are listed below ZZ2 U U z U z values are taken from the Radio Sounding or analitycal profile given in the free formatted part of the PRE IDEAT nam file Y 7 U F Y U Z The U z values are build in the same way as the ZZ
199. iptrc file you may do cp MESONH procedures scriptrc SIMUL scriptrc They may be again modified when calling a procedure with an explicit argument list Script VAR VALUE script will give you the list of the VAR accepted by the script They may also be changed during the full screen initialization step when ENVIRONMENT is set to INTERACTIVE Input files can be used by some scripts prepsource e g in SILENTINTERACTIVE These input files have the general name in script and do not concern the input pa rameters themselves but they contain input that is read by script Thus whereas in INTERACTIVE script would ask the user for information in SILENTINTERACTIVE it reads the data from file inscript 16 CHAPTER 2 THE MESONH PROCEDURES EXCEPT PREPSOURCE 2 2 complete Meso NH work session For the following we recall that host stands for the machine on which the preparative scripts are running and remote host stands for the machine on which the Meso NH job is actually executed 2 2 1 Source extraction and compilation prepsource This part will be described later in chapter 7 because it is only useful when you do not want to use the MESONH model as it stands 2 2 2 Running the model Preparing an initial file or Computing diagnoctic fields prepmodel procedure prepmodel prepares a Meso NH job to run the compiled code with a given main program For most of them PREP IDEAL CASE PRE
200. is configuration allows to create a master library named something a this is useful only for the Master adminis trator BIBMASTER 0 BIBBUGFIX bugfix_name and BIBUSER something a Incorrect case BIBMASTER 0 BIBBUGFIX bugfix name and BIBUSER 0 Incorrect case prepsource performs the following actions 1 it takes out the source files that are asked by the user on LOCAL HOST for that prepsource calls nh takeout and the user must give to the procedure the name of the sccs libraries and the VID An infinite loop is running asking you if you want to take out a new file from a sccs library or to exit out of this loop It should be noted that a whole set of sccs libraries present in the directory with the same VID can be processed all together by using the global mode of the procedure nh takeout If the source file which contains the results of the nh takeout procedure already exists on your local directory prepsource will ask you whether it can overwrite it or not 2 the source is split into its elementary Fortran 90 modules PROGRAM SUBROUTINE FUNCTION MODULE 3 Part of the model code is duplicated when more than one model is used grid nesting active or NSOURCE 2 or more In this case the code is repeated NSOURCE times maximum value for NSOURCE is 8 only changing the model number from one version to the other 4 it also prepares the sequence of make commands in the right order for Fortran 90 first modules then sour
201. ist associated to a given prognostic variable 4 2 30 Namelist NAM BU RU budget for U Fortran type default value LBU_RU NASSEU NNESTU NADVXU NADVYU NADVZU NFRCU NNUDU NCURVU NCORU NDIFU NRELU NVTURBU NHTURBU NPRESU budget switch time filter Asselin nesting advection along x advection along y advection along z forcing nudging curvature terms Coriolis term numerical diffusion relaxation vert turb diffusion hori turb diffusion pressure term logical integer integer integer integer integer integer integer integer integer integer integer integer integer integer nj gt B 42 THE INPUT EXSEGS N NAM FILE 4 2 31 LBU RV NASSEV NNESTV NADVXV NADVYV NADVZV NFRCV NNUDV NCURVV NCORV NDIFV NRELV NHTURBV NVTURBV NPRESV budget switch time filter Asselin nesting advection along x advection along y advection along z forcing nudging curvature terms Coriolis term numerical diffusion relaxation hori turb diffusion vert turb diffusion pressure term Namelist NAM BU RV budget for V logical integer integer integer integer integer integer integer integer integer integer integer integer integer integer 4 2 32 Namelist NAM BU RW budget for W LBU RW NASSEW NNESTW NADVXW NADVYW NADVZW NFRCW NNUDW NCURVW NCORW NGRAVW NDIFW NRELW NPRESW NHTURBW NVTURBW budget switch time
202. it gives a Fortran execution error The informations are written in the following form LUNITn FMFILE 1 ae CONFn LUSERV T LUSERC F LUSERR F LUSERI F LUSERS F LUSERG F LUSERH F NSV 0 amp NAM_LUNITn is the name of the first namelist of this file the character indicates the end of the list of informations The parameters are valuated by VAR VALUE and these prescriptions are separated one from each others by a comma and a blank character Note that you can use more than one line to give one namelist but in this case it is imperative to let a blank character at the end of each line The Meso NH user does not need to prescribe all the parameters of one namelist the missing informations are taken equal to the default values written in the fortran code For example the second namelist in the previous example can be written as amp NAM_CONFn LUSERV T because the other variables of amp NAM CONFn are set to the default values In order to clearly separate what can be modified for a given step of a numerical experiment we affect a different namelist file name for each step see tab 2 1 e To PREpare an Meso NH file containing PhysioGraphical Datas gt file PRE_PGD1 nam e To PREpare an Meso NH file with PhysioGraphical Datas in conformity gt file PRE NEST PGD1 nam e To PREpare an ZOOMed Meso NH file with PhysioGraphical Datas gt file PRE_ZOOMI1 nam 213 214 CHAP
203. itch time filter Asselin nesting total advection advection along x advection along y advection along z forcing numerical diffusion relaxation negative correction sedimentation spontaneous freezing deposition on snow riming of cloud water conversion melting conversion freezing wet growth of graupel dry growth of graupel graupel melting wet growth of hail ICE4 logical integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer gt E w C OQ C SS OO Or Oc coc ou E 94 CHAPTER 4 PERFORM A MESONH SIMULATION 4 2 41 Namelist NAM BU budget for hail Fortran tpe default value LBU RRH NASSERH NNESTRH NADVRH NADVXRH NADVYRH NADVZRH NFRCRH NDIFRH NRELRH NNEGARH NSEDIRH NWETGRH NWETHRH NHMLTRH LBU RSV NASSESV NNESTSV NADVSV NADVXSV NADVYSV NADVZSV NFRCSV NDIFSV NRELSV NDCONVSV NHTURBSV NVTURBSV NCHEMSV budget switch time filter Asselin nesting total advection advection along x advection along y advection along z forcing numerical diffusion relaxation negative correction sedimentation wet growth of graupel wet growth of hail hail melting budget switch time filter Asselin nesting total advection advection along x advection along y advection along z forcing nume
204. ition of graupel riming of cloud accretion of rain conversion freezing wet growth of graupel dry growth of graupel graupel melting ice melting bergeron findeisen cond deposition ice wet growth of hail melting of hail logical integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer gt w C CO O S OQ OQ CIC OOO OS SO oc oO c Q OO oo amp OO o o 42 THE INPUT EXSEGS N NAM FILE 4 2 34 Namelist NAM BU RTKE budget for TKE Fortran type default vale LBU_RTKE budget switch logical NASSETKE time filter Asselin integer NADVTKE total advection integer NADVXTKE advection along x integer NADVYTKE advection along y integer NADVZTKE advection along z integer gt ij NFRCTKE forcing integer NDIFTKE numerical diffusion integer NRELTKE relaxation integer NDPTKE dynamic production integer NTPTKE thermal production integer NDISSTKE dissipation of TKE integer NTRTKE turbulent transport integer SO OO OF COO Oo oo 4 2 35 Namelist NAM_BU_RRV budget for vapor Fortran type default value LBU_RRV budget switch logical NASSERV time filter Asselin integer NNESTRV nesting integ
205. itude km CDADATMFILE if LBOGUSSING TRUE name of the dad of HATMFILE CDADBOGFILE if LBOGUSSING TRUE name of the dad of CINIFILE Pro gram will check that CDADATMFILE and CDADBOGFILE have the same charac teristics before replacing the dad name of CINIFILE by CDADBOGFILE instead of CDADATMFILE CDADBOGFILE must exist before running the prep_real_case job 5 namelist NAM AERO CONE defines aerosol initialization LDUST logical LORILAM logical XINIRADIUSI real XINIRADIUSJ real CRGUNIT character LEN 4 XINISIGI real XINISIGJ real XNOIMIN real XNOJMIN real 166 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES LDUST to switch on the initialization of dust aerosols LORILAM to switch on the initialization of chemical aerosols in case of presence of large scale fields in the atmospheric file from Mocage model XINIRADIUSI to switch on the initialization of mean radius mode I aitken mode of the distribution in micrometers XINIRADIUSJ to switch on the initialization of mean radius mode J accumulation mode of the distribution in micrometers CRGUNIT type of mean radius given in namelist Default is for a mass spectral distribution XINIRADIUSI and XINIRADIUSJ has been converted into a mean radius in number If not default you need to give the mean radius for a number spectral distribution no conversion XINISIGI value of standard deviation for mode I Aitken mode XINISIGJ
206. l the super computer or its storage machine supc archiv 5 6 EXTRACTION OF M T O FRANCE FILES EXTRACTARPEGE 157 OUTDIR directory of the output GRIB file starting at HOME for archiv machine home starting at HOME for super computer storage work starting at workdir for super computer storage If you choose to extract MOCAGE outputs you may indicate the list of the chemical species you want to retrieve in a file named MOC1 nam If not present in your SIMUL directory the following list of default values from the RACM scheme is taken into account and copied in SIMUL MOC1 nam for next use BEGIN CONVMOC2GRIB RACM species without SULF ORA1 ORA2 70 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 HB HH RP nn gn HB HH H HH HB HH NUMBER OF OPTIONAL GRIB VARIABLES 0 3 H 20 2 NO NO 2 NO 3 N 20 5 HONO HNO 3 HNO_4 DMS S0 2 CO 0 1 OH HO 2 CH 4 ETH HC3 HC5 HC8 ETE OLT OLI DIEN ISO API LIM TOL XYL CSL HCHO ALD KET GLY MGLY DCB MACR UDD HKET ONIT PAN 158 222 1 TPAN 223 1 0 1 224 1 0 2 225 1 PAA 226 1 M02 227 1 ETHP 228 1 HC3P 229 1 HC5P 230 1 HC8P 231
207. l velocity WALL for wall boundary condition zero normal velocity e CLBCY array containing 2 elements they represent the type of lateral boundary con dition at the left and right boundaries along y CLBCY 1 and CLBCY 2 respectively The possible values are identical to those for CLBCX e XCPHASE imposed phase velocity of the outgoing gravity waves This phase velocity can be used in the Sommerfeld equation which gives the temporal evolution of the normal velosity at the open lateral boundary 4 2 15 Namelist LUNIT file names Fortran type default value CINIFILE 28 characters INIFILE CCPLFILE array 28 characters JPCPLFILEMAX NONE It contains the names of the different files used for the initialization of the model n They are included in the declarative module MODD LUNITn e CINIFILE name of the initial FM file which contains the field values used as initial state in the present MESONH numerical simulation e CCPLFILE name of the FM files which contains the field values used for the coupling of the outermost MESONH model No more than JPCPLFILEMAX 24 for the present version files can be used in a simulation These CCPLFILE files name are only meaningful for the outermost model which finds its boundary conditions from a previously executed run of Meso NH or another model No constaint are imposed to the coupling file names but they must be tempo rally ordered 68 CHAPTER 4 PE
208. le prepmodelrc stored on SIMUL and can be modified in an interactive session We now give their list 2 2 A COMPLETE MESO NH WORK SESSION 17 e MKFNAME file which contains the makefile e BIBMASTER master library s name e BIBBUGFIX bugfix library s name or 0 none e BIBUSER user library s name starting at HOME or 0 none e DEBUG option of the run run normal execution or debug for debugging option e LOAD OPT loading options added to the default ones they must be between double quote example Inag m e MAINPROG one of the MesoNH programs PREP IDEAL CASE PREP_REAL_CASE PREP PGD PREP_NEST_PGD ZOMM PGD MODEL DIAG SPAWNING or another name to run any program developped in the Mesonh environment e NAMELISTFILE is set to default if MAINPROG is one of the MesoNH programs see tab 2 1 else it must be set to the filename which contains the namelists for the MAINPROG program e LISTGET is set to default if MAINPROG is one of the MesoNH programs else it must be set to the list of the variables which are extracted from the NAMELIST FILE to find the name s of the input FM files e INHOST input mesonh files are on execution remote host or storage file machine supc archiv e INDIR directory of the input mesonh files starting at HOME 0 HOME for storage file machine home starting at HOME for remote host storage work starting at workdir for remote host storage e INLOGI
209. lence scheme used to parameterize the transfers from unresolved scales to resolved scales CTURB no turbulence scheme 42 THE INPUT EXSEGS N NAM FILE 69 CTURB TKEL turbulence scheme with a one and a half order closure i e prognostic turbulent kinetic energy TKE and diagnostic mixing length e CRAD gives the type of radiative transfer scheme used to parameterize the effects of the solar and infrared radiations CRAD NONE then the downward surface fluxes are set to zero CRAD TOPA the solar flux is equal to the one at TOP of Atmosphere The infra red flux is equal to 300 Wm7 CRAD FIXE then the daily evolutions of the downward surface fluxes are pre scribed The temporal evolution is done in the routine PHYS PARAMn by fixing the hourly value of the infrared and solar fluxes and can be modified for personnal application CRAD ECMW the ECMWF radiation scheme code is used e CCLOUD gives the microphysical scheme used to parameterize the different water phases transformations CCLOUD NONE no microphysical scheme You nevertheless may use vapor if desired LUSERV TRUE or FALSE CCLOUD REVE only the saturation adjustment is performed in order to create cloud water if saturation is achievd This liquid water is never transformed in rain water CCLOUD KESS a warm Kessler microphysical scheme is employed It allows the t
210. lied to 0 e XAMPLITH Perturbation amplitude maximum for 0 e XAMPLIRV Perturbation amplitude maximum for ry e XAMPLIUV Perturbation amplitude maximum for U and V e XAMPLIWH Perturbation amplitude maximum for the normalized white noise e NKWH Upper level of the layer starting from the ground where the white noise is applied e LSET_RHU Conservation of the relative humidity TRUE the relative humidity is conserved in the 0 perturbation FALSE the perturbation is computed with the XAMPLIRV amplitude e XCENTERZ Height of the maximum of the 0 perturbation e XRADX X radius of the perturbation e XRADY Y radius of the perturbation e XRADZ Z radius of the perturbation 30 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE 3 2 14 Namelist NAM REAL PGD file switches Fortran type default value CPGD_FILE characters LREAD ZS logical FALSE LREAD GROUND PARAM logical FALSE e CPGD FILE name of the physiographic data file containing the ground data fields The file must be generated by the program For a purely ideal case the CPGD FILE variable may be deleted from the namelist or set to its default value The horizontal grid will be read in the PGD file and therefore the mesh increments XDELTAX and XDELTAY are no more used e LREAD GROUND PARAM switch to use or not the surface cover types COVERnnn and all other physiographic fields except orographic ones read in the P
211. ll the physiographic data necessary to run the MESO NH model with interactive surface schemes for vegetation and town 2 extractecmwf or extractarpege it extracts the surface and altitude fields for one date respectively for ECMWF archive ECMWF forecast model or METEO FRANCE opera tional archive ARPEGE and ALADIN models In both cases the fields are written in GRIB format file on the gaussian grid The extraction must be done separately for each date and time for the initial file and each of the coupling file 3 PREP REAL CASE this MESO NH program performs the change of orography and vertical grid and writes the MESO NH file which will be used either for the beginning of the simulation or for coupling 135 136 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES d gt horizontal interpolation gt vertical interpolation data files ecoclimap orography clay sand files any domain allowed u v T q sp i PGD fields Ts spec zs surface physiographic fields 7f 2 A GRIB file from ECMWF ARPEGE ALADIN or MOCAGE MESONH Physiographic data file small domain fine grid fine resolution lt u v T q sp fine PGD fields Ts interpolated spec zs 222 large scale vertical grid 1 gt u V W 9 T p fine scale zs fine PGD fields Ts MESONH ini
212. locity levels You must start from the ground level K 2 to the model top K KMAX 2 thus you only have to enter KMAX 1 values For instance for the PRE_IDEA1 nam described above it leads to 17 levels 3 3 2 Radiosounding case keyword RSOU The radiosounding data are written in the free format part of PRE_IDEA1 nam file where the altitude variable is e the pressure case KIND STANDARD or PUVTHVMR or PUVTHVHU or PUVTHDHU or PUVTHDMR real in Pascal e the height in case ZUVTHVMR or ZUVTHVHU or ZUVTHDMR or ZUVTHLMR real in meters The first wind variable is e the wind direction in case KIND STANDARD real in degrees 3 3 FREE FORMAT PART 37 the zonal wind in cases KIND PUVTHVMR or PUVTHDMR or ZUVTHDMR or ZUVTHLMR or ZUVTHVHU or PUVTHDHU or ZUVTHVMR or PUVTHVHU real in m s The second wind variable is the force direction in case KIND STANDARD real in m s the meridian wind in cases KIND PUVTHVMR or PUVTHDMR or ZUVTHDMR or ZUVTHLMH or ZUVTHVHU or PUVTHDHU or ZUVTHVMR or PUVTHVHU real in m s The temperature variable is the temperature in case KIND STANDARD real in Kelvin the virtual potential temperature in cases KIND PUVTHVMR or PUVTHVHU or ZUVTHVMN ZUVTHVHU real in Kelvin the dry potential temperature in cases KIND PUVTHDMR or PUVTHDHU or ZUVTHDMR r
213. lt gt lt gt Sia Shi So Org lt p gt Org lt 000 gt zz lt tol gt lt gt Ta 0x8 TP TR SBGT The terms of spectral transport from resolved to subgrid motions is SBTG in the equation of lt E gt sink and ADVR and DPR in the equation of e gt sources One should note that ADVR DPR SGBT 112 CHAPTER 4 PERFORM A MESONH SIMULATION Note in case of grinesting In case of 2way gridnesting the subgrid scheme is not alone to influence the resolved motions due to subgrid scale Part of the job is done by the averaged of the smaller scale models The terms of spectral transport from resolved to subgrid motions are then both SBTG and NEST in the equation of lt E gt sinks Therefore ADVR DPR SGBT NEST Where DPR is the dynamical production that should produce the subgrid scale model to equilibrate the sink at resolved scale 44 BUDGET OF RESOLVED SUBGRID TURBULENT QUANTITIES 113 field notation in processus dim comments I poe gt e E e lt ue gt lt u gt UKE SBG DP M 2 1 mean gradients Ee e men ee fluctuations lt gt 4 lt gt U KE SBG ADVM z t advection pe resolved flow EA Za a FORC z t advection by large scale lt 5 gt subgrid turbulent transport subgrid turbulent transport transport EI lt ul se gt BU_K SBG PRES subgrid pressure correlation out
214. lustrative example of use of procedures 221 A Example of source manipulation 221 A 2 Example of the prepsource script 222 Example of the prepmodel script 225 A 4 Example of the tosupc script 226 Chapter 1 Introduction MESONH is an atmospheric simulation model which can be run in very different conditions Its capacities are completely described in the scientific documentation of the model where the model equations and the whole parametrizations set are given In this book is given the necessary information to perform numerical experiments using the MESONH atmospheric system from the generation of the initial fields to the visualization of the results It will help the user to correctly manipulate all the UNIX scripts source file management to prepare or exploit a numerical experiment made by this model It is intended to give a clear presentation of how to work with the model for a new user and to be a reference guide for a more advanced one For this reason you will find illustrative examples in each chapter of this book in order to have practical applications From a technical point of view no special knowledge are required to perform a simulation It is useful to specify the main characteristics of the model e The sources are written in standard Fortran 90 Metcalf and Reid 1993 e The source file management
215. ly once to compute the initial file for the model 3 Do not change the vertical grid 13 MESONH here is the nested run with model 1 and model 3 CHEESE ve SITIS 2 domain of the PGD file for models 2 and 3 Figure 5 6 Exemple of a grid nesting simulation with 3 nested models the domain of the 2 finest models has the same resolution and a common zone to follow atmospheric system 144 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES 5 2 Creation of MESO NH physiographic data file PREP_PGD physiographic fields are averaged or interpolated on the MESO NH physiographic grid by the program PREP_PGD They are stored in FM file called PGD file but with fewer ele ments than a MESO NH file With the physiographic 2D fields the geographic and grid data are written in this file Note that a MESO NH simulation runs on the grid defined here in PREP PGD 5 2 1 Execution PREP PGD is run with the MESO NH procedure prepmodel In the file prepmodelrc the input host directories and login control variables refer to the input data files and if requested to the FM file YINIFILE The other control variables to initialize specifically in this file are e MAINPROG PREP PGD e NAMELISTFILE default e LISTGET default namelist file PRE PGD1 nam is also needed 5 2 2 Presentation of input files For the version 1 of the externalized surface PREP_PGD needs a minimum of four files to build relatively accurate physiogr
216. me parts of the domain Such records are obtained by asking for temporal series 4 2 27 budgets 4 2 29 aircraft or balloon 4 2 25 profiler or station 4 2 26 LES diagnostics 4 3 See the documentation CREATION et EXPLOITATION de FICHIERS DIACHRONIQUES J Duron oct 2001 for more details 8 2 3 The physiographic file It is a bidimensional MesoNH file with contains surface datas as orography vegetation classes chemical emissions etc See the documentation THE EXTERNALIZED SURFACE USER S GUIDE for more details 8 3 References J Clochard 1989 Logiciel de Fichiers Index s Direction de la Mtorologie Nationale Note de travail ARPEGE n 12 J Clochard 1991 Logiciel de Fichiers Index s Direction de la Mtorologie Nationale Technical report D Gazen 1999 Parallel IO routines Man page on Meso NH web site C Fischer 1994 File structure and content in the Meso NH model Meso NH note Appendix A An illustrative example of use of procedures A 1 Example of source manipulation One classical way to obtain your user library is to copy the Master library in your own directory but you can also create a new library containing only supplementary compared to the Master library sources For example a user has a source file source f90 which he wants to manage with the Meso NH macros First he creates his RCS library with the name MASTER for the first release of this library nh create source f90
217. melist NBERFITH in namelist NBUIH in namelist NBUIL in namelist NBUJH in namelist NBUJL in namelist NBUKH in namelist NBUKL in namelist NBUMASK in namelist NBUMOD in namelist NCDEPIRI in namelist NCDEPIRV in namelist NCDEPITH in namelist NCFRZRG in namelist NCFRZRI 235 NAM BU RTH 85 NAM BU RTKE 86 NAM BU RU 83 NAM BU RV 84 NAM BU RW 84 NAM BU RRC 87 NAM BU RRR 88 NAM BU RRI 89 NAM BU RRS 90 NAM BU RRC 87 NAM BU RRI 89 NAM BU RTH 85 NAM BUDGET 82 NAM BUDGET 82 NAM BUDGET 82 NAM BUDGET 82 NAM BUDGET 82 NAM BUDGET 81 NAM BUDGET 81 NAM BUDGET 81 NAM BU RRI 89 NAM BU RRV 86 NAM BU RTH 85 NAM BU RRG 90 236 in namelist NCFRZRR in namelist NCFRZTH in namelist NCHEMSV in namelist NCMELRG in namelist NCMELRS in namelist NCONDRC in namelist NCONDRV in namelist NCONDTH in namelist NCORU in namelist NCORV in namelist NCORW in namelist NCURVU in namelist NCURVV in namelist NCURVW in namelist NDAD in namelist NDCONVRC in namelist NDCONVRI in namelist NDCONVRV in namelist NDCONVSV in namelist NDCONVTH in namelist NDEPGRG in namelist NDEPGRV NCH SUBSTEPS in namelist NAM CH MNHCn 75 NCH VEC LENGTH in namelist NAM CH MNHCn 75 NAM BU RRI 89 NAM BU RHRR 88 NAM BU RTH 85 NAM BU RSV 91 NAM BU RRG 90 NAM BU RRS 90 NAM BU RRV 87 NAM BU RRV 86 NAM BU RTH 85 NA
218. melist NFRCTH in namelist NFRCTKE in namelist NFRCU in namelist NFRCV in namelist NFRCW in namelist NGMLTRG in namelist NGMLTRR in namelist NGMLTTH in namelist NGRAVW NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU in namelist NAM BU RRI 89 namelist NAM BU RRR 88 in namelist NAM BU RRS 90 in namelist NAM BU 85 in namelist NAM NESTING 57 in namelist NAM PARAM CONVECTn 68 NAM GRIDH PRE 27 RRC 87 RRG 90 RRH 91 RRI 89 RRR 88 RRS 90 RRV 86 RSV 91 RTH 85 RTKE 86 RU 83 RV 84 RW 84 RRG 90 RRR 88 RTH 85 in namelist NHALO in namelist NHENURC in namelist NHENURI in namelist NHENURV in namelist NHENUTH in namelist NHMLTRH in namelist NHMLTRR in namelist NHMLTTH in namelist NHONRC in namelist NHONRI in namelist NHONTH in namelist NHTURBRC in namelist NHTURBRI in namelist NHTURBRV in namelist NHTURBSV in namelist NHTURBTH in namelist NHTURBU in namelist NHTURBV in namelist NHTURBW in namelist NICE NILOC NIMAX NIMLRI NAM BU NAM CO NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU NAM BU RW 84
219. ments microphysical schemes lagrangian trajectory or the chemistry options are treated automatically by the model and should not be counted here 42 THE INPUT EXSEGS N NAM FILE 65 4 2 13 Namelist NAM DYNn parameters for the dynamics of model n Fortran type default value XTSTEP real 60 CPRESOPT 4 characters CRESP NITR integer 4 LITRADJ logical FALSE XRELAX real 1 LHORELAX UVWTH logical FALSE LHORELAX_RV logical FALSE LHORELAX_RC logical FALSE LHORELAX_RR logical FALSE LHORELAX_RI logical FALSE LHORELAX_RS logical FALSE LHORELAX_RG logical FALSE LHORELAX_RH logical FALSE LHORELAX_TKE logical FALSE LHORELAX_SV array logical FALSE LHORELAX SVC2R2 logical FALSE LHORELAX SVCIR3 logical FALSE LHORELAX SVLG logical FALSE LHORELAX SVCHEM logical FALSE LHORELAX SVDST logical FALSE LHORELAX SVAER logical FALSE LVE RELAX logical FALSE NRIMX integer 1 NRIMY integer 1 XRIMKMAX real 1 100 60 XT4DIFF real 1800 It contains the dynamics parameters specific for the model n They are included in the module MODD_DYNn e XTSTEP Time step in seconds If the model is not the DAD model XTSTEP is not taken into account but NDTRATIO in NAM_NESTING e CPRESOPT Pressure solver option choices are implemented in MESONH for the moment see the Scientific documentation for more details RICHA Richardson method preconditionned by the flat cartesian operator CGRAD Generalized pre conditioned gra
220. mpletely NAMLINIFILE CONF PROJ if the grid is a subgrid eventually with higher resolution of an existing one NAM COVER NAM ZS NAMLISBA if you chose to use the ISBA scheme NAM CH EMIS PGD NAM DUMMY PGD Examples of PRE_PGD1 nam file A PREP PGD run where you use the data files provided by the MESO NH team and some files of your own for the dummy fields amp NAM PGDFILE CPGDFILE PGDFILE_1 amp NAM_PGD_GRID CGRID CONF PROJ 146 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES CONF PROJ XLATO 45 XLONO 0 XRPK 0 7 XBETA O CONF PROJ GRID NIMAX 100 NJMAX 100 XLATCEN 42 5 XLONCEN 2 5 XDX 10000 XDY 10000 INIFILE CONF _ PGD SCHEMES CNATURE ISBA CSEA SEAFLX CTOWN TEB CWATER WATFLX COVER YCOVER ecoclimap_v2 YFILETYPE DIRECT 275 YZS gtopo30 YFILETYPE DIRECT ISBA YCLAY clay_fao YCLAYFILETYPE DIRECT YSAND sand_fao YSANDFILETYPE DIRECT XUNIF_RUNOFFB 0 5 DUMMY PGD NDUMMY_NBR 6 CDUMMY_NAME 1 SST_2001062100 CDUMMY AREA 1 gt SEA CDUMMY_ATYPE 1 ART CDUMMY_FILE 1 5Tn2001062100 dat CDUMMY_FILETYPE 1 ASCLLV CDUMMY_NAME 2 SST_2001062200 CDUMMY_AREA 2 gt SEA CDUMMY_ATYPE 2 gt ART CDUMMY_FILE 2 55 2001062200 CDUMMY_FILETYPE 2 ASCLLV CDUMMY_NA
221. n CCH TDISCRETIZATION SPLIT use time splitting input fields for solver arescalar variables at t dt derived from XRSVS CCH TDISCRETIZATION CENTER use centered tendencies input fields for solver are scalar variables at t XSVT CCH TDISCRETIZATION LAGGED use lagged tendencies input fields for solver are scalar variables at t dt XSVM NCH_SUBSTEPS number of steps to be taken by the solver during two time steps of MesoNH the time ste p of the solver is thus equal to 2 XTSTEP NCH_SUBSTEPS LCH_TUV_ONLINE switch to activate online photolysis rates calculations only for 1D simulation If false photolysis rates are pre_calculated as a function of s olar zenith angle and surface albedo and interpolated on the model grid CCH_TUV_LOOKUP name of the lookup table file CCH TUV CLOUDS method for calculating the impact of clouds on UV radiations only for 3 D versi on CCH_TUV_CLOUDS NONE No cloud correction on UV radiations CCH_TUV_CLOUDS CHAWN Cloud correction on UV radiations following Chang et al 1987 XCH_TUV_ALBNEW surface albedo for photolysis rates calculations only for 1 D ver sion For 3 D version albedos are prescribed as a function of the surface characteristics 78 CHAPTER 4 PERFORM A MESONH SIMULATION e XCH_TUV_DOBNEW scaling factor for ozone column dobson e XCH TUV TUPDATE update frequency to refresh photolysis rates e CCH VEC METHOD type of vectorizat
222. n 72 LSIGMAS in namelist NAM TURBn 72 LSLEVE in namelist NAM VER 32 159 LSPAWN SURF in namelist NAM SPAWN SURF 167 LSUBG AUCV in namelist NAM TURBn 72 LSUBG COND in namelist NAM TURBn 72 LTEND THRV FRC in namelist NAM_FRC 56 LTHERMIJ in namelist NAM CH ORILAM 76 LTHINSHELL in namelist NAM VER GRID 31 158 LTURB DIAG in namelist NAM TURBn 72 LTURB FLX in namelist NAM TURBn 72 LUSECHEM in namelist NAM CH MNHCn 74 LUSECI in namelist NAM CONFn 62 LUSERC in namelist NAM CONFn 25 LUSERI in namelist NAM CONFn 25 LUSERV in namelist NAM_CONFn 25 62 LVARSIG in namelist NAM DUST 53 LVARSIGI in namelist NAM CH ORILAM 76 LVARSIGJ INDEX in namelist LVE RELAX in namelist NAM CH ORILAM 76 NAM DYNn 64 LVERT MOTION FRC in namelist LWARM in namelist LWMINMAX in namelist LZDIFFU in namelist N NACCRC in namelist NACCRR in namelist NACCRRR in namelist NACCRS in namelist NACCTH in namelist NADVXRC in namelist NADVXRG in namelist NADVXRH in namelist NADVXRI in namelist NADVXRR in namelist NADVXRS in namelist NADVXRV in namelist NADVXSV in namelist NADVXTH in namelist NADVXTKE in namelist NADVXU in namelist NADVXV in namelist NADVXW NAM FRC 56 NAM_PARAM_ICE 59 NAM SERIES 79 NAM DYN 54 NAM BU RRC 87 NAM BU RHRR 88 NAM BU RHRR 88 NAM BU RRS 90 NAM BU RTH 85 NAM BU RTH 87 NAM BU RRG 90 NAM
223. necessary to complete your binary library on the REMOTE HOST because the sources where no bug were found have been already saved in your binary library This time you can work with ENVIRONMENT SILENTINTERACTIVE Just use the file inprepsource new previously created by inprepsource new inprepsource modify it to keep only the source file s you want to compile again be careful of the syntax used in the file inprepsource each line corresponds to the informations put on the keyboard blank included cf1 90 EXAMPLE modeln f90 EXAMPLE 7 5 HOW TO USE USER S LIBRARY 211 the first and third blanck lines correspond to the carriage return to the question of localisation of RCS sources DIRSRC directory the last one is for the carriage return to the additional compilation options the characters are a separator Thus working first in INTERACTIVE to produce an inprepsource file and then modify it in SILENTINTERACTIVE is an easy way to proceed with prepsource 7 5 How to use a user s library We have seen in the previous chapters how to use the Meso NH model as it stands by collecting the binary objects of the Master and the Bugfix source libraries previously compiled on the remote_host by the Meso NH administrator Once you have compiled your own source files new subroutines created or pre existing ones modified and created on the remote host your User s binary library with their counterpart object files you
224. nh takeout extracts takes out source files and collects them into one file nh register splits the collected source file and registers the modifications for each elementary file undoes a global nh extract nh unregister undoes a global nh register lists the historic of all files All these macros will help you take a look at the sources to modify a source register a modification either at the most basic level file by file or all the libraries present in the directory DIRSRC at the same time Of course the Meso NH user can start from a Fortran source taken from the Master library nh takeout modify this Fortran source with any editor and create a RCS library enclosing its personal version of the Fortran source He also can write from to Z its Fortran source and enclose it in a RCS library It should be noted that all these actions are performed on the local host The next step is to compile on the remote host these new or modified sources and to keep the binaries in binary library in order to avoid repeated compilations This is achieved by the procedure named prepsource presented in the next section 7 3 prepsource procedure This procedure prepares on the localhost a UNIX job which contains all the orders necessary to achieve the compilation and storage of the results on the remote host the Fortran sources have to be ready to be compiled because the prepsource procedure does not allow you to modify these F
225. nt variable SIMUL is not initialized When the job is completed on the remote_host an execution report is send back to the localhost The model output files are disposed where stated in the prepmodelrc file in workdir INOUTFILES for this example It should be noted that it is still the case if the outprepmodel file is executed in an interactive session on the remote host instead of a batch session 134 CHAPTER 4 PERFORM A MESONH SIMULATION Chapter 5 Initialization of MESO NH for real cases 5 1 Overview of the initialization sequences 5 1 1 Initialization for one model run from operationnel models This type of model simulation uses one initial file with all the fields at beginning time of the simulation and a certain number of coupling files to prescribe the lateral boundary conditions to the model at further time steps The coupling files are of the same nature than the initial file Here is presented the way to produce these initial and coupling files This sequence for real cases contains three procedures or programs figures 5 1 and 5 2 In the following a fortran program is written in CAPITAL letters and is run with the MESO NH procedure prepmodel while an unix procedure is written in lower case letters 1 PREP_PGD this MESO NH program computes the physiographic data file called PGD file below At this step you choose the projection resolution and horizontal domain The PGD file contains a
226. nted destruction and to store different versions of the same source It should be noted that you can create one library per Fortran source and collect all these libraries in the same directory called DIRSRC This directory should also be different from the one SIMUL from where the preparative scripts are run and which contains the namelists and rc files This is not obligatory but strongly recommended to simplify the user s life For instance all the sources of the Master library are stored and protected by the same way but on more than one directory because of the large number of involved sources The Meso NH source file manager updates the RCS file by calling the appropriate RCS command The basic RCS file contains the Fortran source code and all the necessary information necessary for RCS to recover the different versions This file is often called the file v since it is suffixed by v The file v also contains the necessary information to recover the RCS version corresponding to a Meso NH version name Before using any Meso NH source management tool you may first initialize the environment variable DIRSRC pointing to the absolute directory pathname that contains either the Master the Bugfix or the User library The UNIX command is DIRSRC user library export DIRSRC If this environment variable is not initialized its value is asked interactively by each Meso NH RCS macro Table 7 1 lists all the Meso NH macros for the source f
227. ntil 20070630 1 ____ 32bit a must have OUTDEST Automatic submit by tosupc for part one tosupcrc with TIME MEM NBP TPN has to be OK LSOUMISAUTO T options for treatment of all FM files and post treatment Ilfiz 21412 conv2dia fm unlfiz all fmmore OUTFILE_TOOLS SUBMIT NEXTJOBS 0 or filenames to submit super_calculator at the end of this job tosupcrc must be present on the same directory 0 job one or several scripts between commas which will H be submitted one by one by tosupc 45 AN ILLUSTRATIVE EXAMPLE OF A MESONH SIMULATION 129 the pathnames are relative to the current SIMUL directory or absolute SUBMIT NEXTJOBS 0 LOCAL NEXTJOBS 0 or filenames to execute workstation jobs at the end of this job list of jobs executed on local workstation at the end of this job second argument list of sh command arguments which will run on your workstation with at procedure the pathnames are relative to the current SIMUL directory or abso O no job LOCAL_NEXTJOBS dirlocal tracei the workstation job will execute cd dirlocal sh tracel LOCAL NEXTJOBS 0 name of your personal makefile if you don t have one general makefile will be executed on the remote machine according its 08 MKFNAME make_mnh user binary library 0 if none will be searched from your HOME BIBUSER O reference binary library BIB
228. ocal host e g a trace session after computation of diagnostics 2 2 3 Send your job for execution tosupc Output scripts prepared before outprepsource or outprepprog with shell commands have to be executed on the remote host Here we describe what tools are used to submit and execute the output script The procedure tosupc allows either to send the job to the remote host and to put it into its wait queue or to run the job on the local host Thus tosupc runs a job in batch on the remote host or in batch or interactive on the local host depending on the control parameters described below Another way to work is to execute the job interactively on the remote job It can be put on the remote host by ftp or rcp and then executed interactively if it is possible to 2 2 A COMPLETE MESO NH WORK SESSION 19 do it it is safier to work on the tmp of the remote host Just execute on this machine the following order outscript Here are the control parameters for tosupc file tosupcrc e JOBFILE filename which contains the script to be run e JOBNAME job name on the remote host JOBFILE by default e TIME time in seconds class e MEM memory with unit default is mbytes except mwords for Cray e TPN number of processors on each node e NBP mono or multi node job e MAIL is sent at the beginning or the end or both of the job tosupc calls another UNIX procedure which effectively transfers the ouput script file to
229. odel must be selected even if the grid nesting is active e NBUMASK Number of masks used to select the budgets areas the case CRUTYPE 5 e XBULEN Duration in seconds on which the different source terms of all the budget temporally averaged the minimum value is 2 XTSTEP e XBUWRI Duration in seconds between successive writings in the diachronic file of the budget storage arrays CBUTYPE MASK e NBUKL value of the model level K for the bottom of the budget box in the case of a cartesian box CBUTYPE CART 42 THE INPUT EXSEGS N NAM FILE 85 e NBUKRH value of the model level K for the top of the budget box in the case of a cartesian box CBUTYPE CART Inside the budget box NBUKL lt K lt NBUKH e NBUJL value of the model level J for the left side of the budget box in the case of a cartesian box CBUTYPE CART e NBUJH value of the model level J for the right side of the budget box in the case of a cartesian box CBUTYPE CART Inside the budget box NBUJL lt lt NBUJH e NBUIL value of the model level I for the left side of the budget box in the case of a cartesian box CBUTYPE CART e NBUIH value of the model level I for the right side of the budget box in the case of a cartesian box CBUTYPE CART Inside the budget box NBUIL lt J lt NBUIH e LBU_KCP Logical switch to average or not in the K direction all the budget terms for any CBUTYPE
230. odelrc file PREP SURFEX is run with the procedure prepmodel The input and output host directory and login control variables refer to all the input and output files and not only to input or output FM files The other control variables to initialize specifically are e MAINPROG PREP_SURFEX e NAMELISTFILE default e LISTGET default These variables contain the names of the input atmospheric file and Physiographic data FM file respectively 5 9 2 physiographic data file This is a FM file but with fewer elements than a MESO NH file It contains the physiographic 2D fields geographic and grid data are stored on this file This file is created by the program PREP PGD It is possible to use a complete MESO NH file since it also contains the physiographic fields If one wants only modify the vertical grid of MESO NH file without any change on the orography one can specify it both as atmospheric file and physiographic data file 5 9 IF YOU WANT SURFACE FIELDS ONLY PREP_SURFEX 173 It contains e the definition of the projection the horizontal domain and the horizontal grid e the physiographic fields 5 9 3 The input file Both GRIB and FM file are self explanatory The physiographic data stored in it will not be saved on the output file 5 9 4 The file PRE REAL1 nam This file contains namelists with the directives to run PREP SURFEX The namelists contain the names of the files The order of namelist is fr
231. om NAMELISTFILE LISTGET default j o input files location of the input FM files on execution machine or storage machine HOME INDIR supc archiv INHOST supc to get the input FM files directory name starting at HOME if it begins with home at workdir if it begins with work indicate one or several directories between double quotes INDIR work INOUTFILES user name on which the get will be executed O LOGNAME oct INLOGIN 0 Overview of archiv case INHOST archiv INDIR mxxx mxxx007 CHEMIN mgrp999 NEXTDOOR INLOGIN 0 OR INHOST archiv INDIR NEXTDOOR INLOGIN mrgp999 output files storage of the output FM files on execution machine or storage machine supc archiv RMACH aerosv2 OUTHOST supc to put the output FM files directory name starting at HOME if it begins with home at workdir if it begins with work OUTDIR work INOUTFILES global variables used in prepmodel DEBUGSCRIPT ON OFF ENVIRONMENT SILENTINTERACTIVE INTERACTIVE FILE tosupcrc bin sh default input variables for tosupc default values for the 05 parameters Asencio 13 12 94 specific variables filename which contains the script to be run on remote host 3 5 AN ILLUSTRATIVE EXAMPLE OF PREP IDEAL CASE AT JOBFILE 11 0 jobname JOBNAME 12 basename JOBFILE tosupc will not
232. onents LWIND_ZM suffixed by _ZM for zonal and meridien components FoRCing variables at instant n LFORCING in YINIFILE des UFRCn VFRCn WFRCn 1D m s LVAR_FRC THFRCn RVFRCn TENDTHFRCn TENDRVFRCn 1D s GXTHFRCn GYRVFRCn 1D PGROUNDFRCn 0D Convective scheme CAPE 2D Convective Available Potentiel Energy J kg NCONV KF 0 1 2 CLTOPCONV CLBASCONV 2D top and base of conv clouds km 52 DTHCONV 3D Convective tendency for potential temperature K s NCONV KF 1 2 DRVCONV DRCCONV DRICONV for vapor cloud ice s Ky DSVCONVnnn P for scalar variable n s i UMFCONV DMFCONV 3D Updraft and Downdraft Mass Flux m kg s NCONV KF 2 PRLFLXCONV PRSFLXCONV 3D Liquid and Solid PRecipitation FLuX m s 5 6 3 VARIABLES AVAILABLE IN THE OUTPUT DIACHRONIC FILE 183 Turbulent scheme TKEM 3D Turbulent Kinetic Energy m s LVAR TURB SIGS 3D from turbulence scheme kg kg Iyl SROM 3D Normalized 2nd_order moment EUR kg kg c BL DEPTH 3D Boundary Layer Depth if CTOM TMO6 m ATHETA AMOIST 3D m RED_TH1 RED R1 RED2 TH3 RED2 R3 RED2_THR3 3D Redelsperger numbers DP TP TR DISS 3D m s dyn therm production transport and dissipation of TKE PHI3 PSI3 PSI SV n LM THLM RNPM 3D m _ _ m K kg kg e 1d scheme turbulent fluxes THW FLX RCONSW FLX RCW FLX 3D m s kg kg m s THL VVAR THLRCONS VCOR RTOT
233. onnees UM 22378 mots position 2822 a 25199 25 eme article de donnees VM 22378 mots position 25200 a 47577 26 eme article de donnees WM 22378 mots position 47578 a 69955 27 eme article de donnees THM 22378 mots position 69956 a 92333 28 eme article de donnees TKEM 22378 mots position 92334 a 114711 29 eme article de donnees PABSM 22378 mots position 114712 a 137089 30 eme article de donnees RVM 22378 mots position 137090 a 159467 31 eme article de donnees RCM 22378 mots position 159468 a 181845 32 eme article de donnees RRM 22378 mots position 181846 a 204223 33 eme article de donnees RIM 22378 mots position 204224 a 226601 34 eme article de donnees RSM 22378 mots position 226602 a 248979 35 eme article de donnees RGM 22378 mots position 248980 a 271357 36 eme article de donnees LSUM 22378 mots position 271358 293735 37 eme article de donnees LSVM 22378 mots position 293736 a 316113 38 eme article de donnees LSWM 22378 mots position 316114 a 338491 39 eme article de donnees LSTHM 22378 mots position 338492 a 360869 40 eme article de donnees LSRVM 22378 mots position 360870 a 383247 41 eme article de donnees LSXTKEM s 298 mots position 383248 a 383545 42 eme article de donnees LSYTKEM 14926 mots position 383546 a 398471 43 eme article de donnees LSXRCM n 298 mots position 398472 a 398769 44 eme article de donnees LSYRCM 14926 mots posi
234. or rain water logical integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer integer Hj gt img cc oco Co Co 91 92 CHAPTER 4 PERFORM A MESONH SIMULATION 4 2 38 Namelist NAM BU budget for non precipitating ice Fortran type default value LBU_RRI budget switch logical NASSERI time filter Asselin integer NNESTRI nesting integer NADVRI total advection integer NADVXRI advection along x integer NADVYRI advection along y integer NADVZRI advection along z integer NFRCRI forcing integer NDIFRI numerical diffusion integer NRELRI relaxation integer NDCONVRI deep convection integer NHTURBRI hori turb diffusion integer NVTURBRI vert turb diffusion integer NNEGARI negative correction integer NSEDIRI sedimentation ICE3 or ICE4 integer NHENURI heterogenous nucleation ICE3 or ICE4 integer NHONRI homogeneous nucleation ICE3 or ICE4 integer NAGGSRI aggregation of snow ICE3 or ICE4 integer NAUTSRI autoconversion of ice ICE3 or ICE4 integer NCFRZRI conversion freezing ICE3 or ICE4 integer NWETGRI wet growth of graupel ICE3 or ICE4 integer NDRYGRI dry growth of graupel ICE3 or ICE4 integer NIMLTRI ice melting ICE3 or ICE4 integer NBERFIRI Bergeron Findeisen gth ICE3 or ICE4 integer NCDEPIRI cond deposition on ice ICE3 or ICE4 integer NW
235. ora replaced by lt Org gt 29 RESRVZ zip o o lt gt sR wtp 7 OOOO SSS lt g gt tp m 7 lt gt RESSVPZ lt a 2 eru gt CS 18 Yle 5 R Sr V e R 3 V v peres s us Sd spore eei reswee Paw Sd 4 3 LES DIAGNOSTICS 105 4 3 6 LES averaged fields LLES SUBGRID TRUE field notation in dim general comments diac file conditions gt we _ wv SBGUV ewes SBGWU ews SBGNV lt gt seme zip For paei wm SBG UTHL ztp __ if nor replaced by wi ww evm ztp __ if nor replaced by cv lt gt SBGOWTHL ztp fno re replaced by cw wW SBGWr V ztp lt gt iore tp n neoon lt a gt SBC TLRT rn by lt 0 gt wr oom n rrisfor total water ems SBGVRr zip n ris for total water ews SBGWRT r risor total water evs fecr few m SSS lt gt SBGURC zip 7 SS ems SBQVRC gt ems eaw lt gt SBQUSV ztpn lt gt 698 ztpn SS Jol cus 1 1 SBG WSV tpn os lt ae id gt a
236. ortran sources These modifications had to be performed with the Meso NH RCS procedures nh create nh extract nh register in order to obtain RCS source libraries which enclose the Fortran sources The main object of prepsource is to create or update a User s binary library on the remote host This library contains binary objects which can be present or not in Master or Bugfix binary libraries the generic term binary object applies either to a main program a subroutine or a module If a binary object is present in both libraries the User s library one will be used when the model is loaded by prepmodel Because the different prognostic Meso NH variables are defined in declarative modules named MODD the compiled form of this module must be available in the Master or User library when a Fortran source is compiled and uses these variables transferred in this routine by a USE association Metcalf and Reed 1993 TThe same problem occurs for the modules which enclose interfaces named MODL If the compiled form of the module is present in the User s library it will be used during the compilations else it will searched in the Master library The last case occurs when the Meso NH user compiles this module in the same job as other Fortran sources which use it In this case prepsource first compiles all the MODL MODD and in a second step the other subroutines Because the binaries of the User s library have priority on the Master s
237. pc default values for the NQS parameters N Asencio 13 12 94 specific variables filename which contains the script to be run on remote host JOBFILE 11 0 jobname JOBNAME 2 basename JOBFILE THHEHHEHHHHHHHHHHHHHHE tosupc will not work unless you correctly fill TIME MEM TPN NBP time in seconds for J2 amp J3 jobs TIME memory ex 2000 2Gb 128Mb tori less than 128Gb per node be careful reduce the memory for multi tasks jobs MEM tasks per node only for prep ideal case run or diag number of used CPUs from 1 to 8 The most important for parallel execution TPN 1 mono or multi node on NEC 1 to 4 and on IBM NBP 1 7 4 AN ILLUSTRATIVE EXAMPLE OF A SOURCE COMPILATION 209 tosupc will not work unless you correctly fill TIME MEM TPN NBP Sending a mail abort beg and or at the end job begend beg end no default is abort no is no mail at all MAIL global variables used tosupc DEBUGSCRIPT 0N OFF ENVIRONMENT SILENTINTERACTIVE SILENTINTERACTIVE BATCH INTERACTIVE You are ready to run the procedure which prepares the compiling orders Stay in the SIMUL directory where the rc files are Work in ENVIRONMENT INTERACTIVE export ENVIRONMENT INTERACTIVE for the first time Enter prepsource 99 The first question asked to you is the localization of the SIMUL directory answer Then a copy of the
238. ple to prepare an initial file for an ideal case experiment The rc files used for the compilation of these sources are prepsourcerc and tosupcrc or mesonhrc We give below a copy of the resulting rc files FILE prepsourcerc bin sh default initialization of environment variables for prepsource c fischer 21 07 94 Se Se eo ae a Se Sa Sa aa Se eS ea Se Se Se eS SS SS instead of outprepsource OUTSCRIPT tori hpce tora 1x32 a must have OUTDEST Automatic submit by tosupc no tosupcrc required LSOUMISAUTO T 208 CHAPTER 7 MODIFY THE FORTRAN SOURCES name for master binary file name user remote binary libraries starting from HOME BIBMASTER DEFBIBMASTER BIBBUGFIX DEFBIBBUGFIX BIBUSER modif a export BIBMASTER BIBBUGFIX BIBUSER other variables debugging option DEBUG run everything else will be debug DEBUG run export DEBUG name of makefile if you ve got a personal one MKFNAME make_mnh export MKFNAME number of duplicate sources spawning multitasking NSOURCE 1 global variables used in prepsource SDEBUGSCRIPT ON ON OFF ENVIRONMENT SILENTINTERACTIVE INTERACTIVE BATCH FILE tosuprc bin sh default input variables for tosu
239. r anelastic system approximated form of the DURran version of the anelastic sytem classical Modified Anelastic Equations but with not any approximation in the momentum equation e NVERB verbosity level 0 for minimum of prints 5 for intermediate level of prints 10 for maximum of prints If CSURF EXTE in namelist NAM GRn PRE NVERB 10 prints two AT pXfiles containing the initialisation of surface scheme variables for each type of surface cover in french or in english e CIDEAL kind of idealized fields CSTN Constant moist Brunt Vaisala frequency case RSOU radiosounding case e CZS orography selector The formulae are given below in the description of the namelist NAM GRIDH PRE 24 CHAPTER 3 PREPARE AN INITIAL MESONH FILE FOR AN IDEAL CASE FLAT constant XHMAX orography zero by default SINE sine shaped orography BELL bell shaped orography DATA discretized orography The data describing the orography are given in the free format part Only the orography corresponding to the computational domain must be provided in free format For 3D orography data are read like if it was a map the first line is the Northern border and the first data is the North West corner with one line per Y axis increment e LBOUSS Switch for a Boussinesq version TRUE The reference anelastic state is cte Oref z 0 and pref cte
240. r compilation on remote machine makefile compilation options default c answer by a carriage return Then prepsource works alone and splits the whole source in its Fortran elements and lists these elements end of silent interactive prepsource loop a new file inprepsource new is now available for future work with prepsource in SILENTINTERACTIVE modi cfl f90 cfl f90 modelni f90 The prepsource step is now over and you will proceed as before with the outprepsource file Send this file in the batch queues of the REMOTE HOST a report listing will be send back to the LOCAL HOST indicating the compilation report You can use a faster way of proceeding prepsource by using global mode for the nh takeout procedure The only difference appears when you answer the following question in prepsource please enter arguments for nh takeout file name version name VID users come stein SOURCE EXAMPLE Then both files cfl f90 and modeln f90 will be taken out from the DIRSRC directory This way of working is very interesting when you have a lot of sccs libraries present on the same directory and all managed with the same VID If compilation errors happen for one or more Fortran sources you must correct the bugs in the sources files by creating new RCS versions nh extract nh register on the LO CAL HOST Then you can compile again these new sources using prepsource Only the com pilation of these sources is
241. r instance It can also need additional files to realize the coupling of the outermost model with a large scale model MESONH user will specify some free parameters of the run by fixing their new values in the NAMELISTs of the file EXSEG n nam When more than one model is present each model needs its own MESONH file to be initialized and its own EXSEG n nam file to fix the free parameters note that a lot of physical free parameters depends on the mesh and therefore vary with the model number The input files are read by the program in order to realize the initialization and the eventual coupling of the MESONH model with a large scale model CEP Arp ge They all have the same standart form described in the first Annexe e a descriptive part It is filled by the copy of the EXSEG n nam file corresponding to the model number in the grid nesting procedure which has been used to perform this segment And at the end of these informations are added the comments describing the fields written in the binary part of the file e a binary part where the fields are written by the LFI subroutines The output files are of two types e synchronous files for a given instant of the run They contain the prognostic fields and eventually additionnal records for supplementary diagnostic fields at the same instant The file name ends by 00n with n gt 0 e a diachronic file for the temporal series of prognostic or diagnostic fields The file name
242. r of the lognormal mode n m DSTSIGAn 3D Dust Standard deviation of the lognormal mode n DSTMSSn 3D Dust Mass concentration of the lognormal mode n ug m DSTBRDNn 2D Dust Burden of the lognormal mode n g m DSTAOD2D 2D Dust Optical Depth _ NRAD 3D 1 DSTAOD3D 3D Dust Optical Depth between two vertical levels _ DSTEXT 3D Dust Extinction 1 km U Chemical aerosol variables LORILAM in YINIFILE des SOAIM 3D Aerosol scalar variable ppbv LCHEMDIAG as defined in ch aer init soa f90 RGAn 3D Aerosol number mean Radius of the lognormal mode n jum RGAMn 3D Aerosol Mass mean Radius of the lognormal mode n 3D Aerosol Number of the lognormal mode n cc SIGAn 3D Aerosol Standard deviation of the lognormal mode n 6 3 VARIABLES AVAILABLE IN THE OUTPUT DIACHRONIC FILE 185 Variables dim meaning unit in DIAG1 nam Chemical variables LUSECHEM in YINIFILE des 03M 3D Chemical scalar variable ppbv as defined in BASIC f90 LCHEMDIAG TRAJectories using Lagrangian tracers X Y 3D X and Y coordinates km LGXM LGYM LGZM 3D Lagrangian tracers coordinates m Yn Zn 3D Lagrangian tracers coordinates at time origin n km THn RVn 3D corresponding Theta and Vapor mixing Ratio K g kg i see Lagrangian analyses documentation on Meso NH web site Books and Guides LLG in YINIFILE des
243. ransformations between 3 classes of water vapor cloud water and rain CCLOUD C2R2 a 2 moment warm microphysical scheme according to Cohard and Pinty 2000 CCLOUD KHKO a 2 moment warm microphysical scheme for LES of Stratocu mulus according to Khairoudinov and Kogan 2000 CCLOUD ICE3 a mixed microphysical scheme including ice snow and graupel 6 classes of hydrometeors CCLOUD corresponding to ICE3 with hail 7 classes of hydrometeors e CDCONV gives the type of deep convection scheme used to parameterize the effects of unresolved convective clouds CDCONV NONE no convection scheme CDCONV KAFR Kain Fritsch Bechtold scheme 70 CHAPTER 4 PERFORM MESONH SIMULATION 4 2 18 Namelist PARAM CONVECTn options for the convective scheme of model n Fortran type default value XDTCONV real MAX 300 0 X TSTEP NICE integer 1 LREFRESH_ALL logical TRUE LCHTRANS logical FALSE LDEEP logical TRUE LSHAL logical FALSE LDOWN logical TRUE LSETTADJ logical FALSE XTADJD real 3600 XTADJS real 10800 LDIAGCONV logical FALSE NENSM integer 0 It contains the options retained for the deep convection scheme used by the model n They are included in the declarative module MODD PARAM CONVECTn XDTCONYV timestep for call of the convection scheme It is limited to 300s NICE flag to include ice proceses in convection scheme 1 yes 0 no ice
244. rced relaxation is enabled when CRE LAX_HEIGHT_TYPE FIXE or minimal height if HGR is taken e CRELAX_HEIGHT_TYPE definition of the height above which a forced relaxation is enabled FIXE means that a forced relaxation is never applied below XRELAX HEIGHT FRC THGR means that a forced relaxation is never applied below the maximal height between XRELAX HEIGHT FRC and the height above which 00 02 is the highest for each column e LTRANS switch to apply a Galilean translation of the domain of simulation TRUE The translation speed of the domain of simulation will be XUTRANS XVTRANS FALSE not active e LPGROUND FRC switch to simulate a time varying ground pressure WARNING THIS FLAG CAN NOT YET ACTIVATED 42 THE INPUT EXSEGS N NAM FILE 59 TRUE means that the ground pressure is updated if it varies in time in the Free Format Forcing Part of the PRE_IDEAI1 nam file FALSE not active 4 2 7 Namelist NAM NESTING grid nesting configuration Fortran type default value NDAD NDTRATIO XWAY array 8 real m 1 array 8 integer array 8 real e NDAD m is the model number of the father of each model m e NDTRATIO m is the ratio between time step of model m and its father NDAD m e XWAY m is the interactive nesting level for model m and its father NDAD m one way interaction 2 two way interaction 3 two way interaction upw
245. rical diffusion relaxation deep convection hori turb diffusion vert turb diffusion chemistry activity logical integer integer integer integer integer integer integer integer integer integer integer integer integer integer logical integer integer integer integer integer integer integer integer integer integer integer integer integer nj gt E cOococccccocococococococccoco 4 2 42 Namelist NAM BU RSV budget for a Scalar Variable Fortran type default value nj gt oooocjcooooqoooo amp 4 2 43 LES budgets Namelist NAM_LES This namelist controls the diagnostics of turbulence especially for Large Eddy Simulations The diagnostics are registered in the diachronic file 000 The list of the diagnostics is given in the following section 42 THE INPUT EXSEGS N NAM FILE 95 Fortran type default value LLES_MEAN logical FALSE LLES_RESOLVED logical FALSE LLES_SUBGRID logical FALSE LLES_UPDRAFT logical FALSE LLES_DOWNDRAFT logical FALSE LLES_SPECTRA logical FALSE NLES_LEVELS integer all levels XLES_HEIGHTS real none NSPECTRA LEVELS integer none XSPECTRA HEIGHTS real none NLES TEMP SERIE I integer none NLES_TEMP_SERIE_J integer none XLES_TEMP_SERIE_Z real none CLES_NORM_TYPE character len 4 NONE CBL HEIGHT DEF character len 3 KE XLES_TEMP_SAMPLING real 60 s if CTURB 3DIM 300 s if CTURB 1DIM
246. rid definition leen 26 3 2 9 Namelist NAM GRIDH PRE horizontal grid definition 27 3 2 10 Namelist NAM GRn PRE soil scheme choice 28 3 2 11 namelist LBCn PRE lateral boundary conditions 28 3 2 12 Namelist NAM LUNITn logical unit names 28 3 2 13 Namelist NAM PERT PRE set analytical perturbations 29 1 CONTENTS 3 2 14 Namelist NAM REAL file switches 3 2 15 Namelist NAM SLEVE smoothed orography for Sleve coordinate 3 2 16 Namelist NAM VER GRID contains vertical grid definition 30 3 2 17 Namelist NAM VPROF PRE variables for CIDEAL CSTN RSOU 32 3 2 18 Namelists of the externalized 3 2 19 Example of namelist part of PREIDEAl nam 3 3 cBreestormat ey Ae ceca fb xS et 3 9 1 Optional Vertical gridi S riens medamne ety RR URS 3 5 2 Radiosounding case 3 3 8 Constant moist Brunt Vaisala 3 3 4 The forced version 3 3 5 Discretized orography 3 4 The output MESONH file 3 5 An illustrative example of PREP IDEAL CASE Perform a MESONH simulation 4 1 What 1 4 2 Th
247. rmal plane coordinates x y and index i j for a FM file with its projection domain and grid This procedure is used as latlon2xy MESONH file name Afterwards the user follows the indications of the program he gives interactively the latitudes and longitudes of the points 176 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES 5 10 4 Conversion from conformal coordinates to sphericalcoordinates xy2latlon This procedures computes the position of points given with coordinates x y on the conformal plane of a FM file with its projection domain and grid onto the earth latitude and longitude This procedure is used as xy2latlon MESONH_file_name Afterwards the user follows the indications of the program he gives interactively the x and y coordinates of the points Chapter 6 Compute diagnostics after a MESO NH simulation 6 1 Presentation After running the model useful quantities can be diagnosed from prognostic variables contained in the FM output files It is done by the program DIAG which computes diagnostic variables If you choose diachronic format for output file you can plot fields with the interactive program diaprog in the super computer or in your workstation Available diagnostics are listed in section 6 3 They can be sorted into 5 main cate gories as seen by the 5 tables You can add your own diagnostics by modifying the source write lfifm1 for diag f90 or write_lfifm1_for_diag_supp f90 6 2
248. rpart on the remote host You will have to compile these new sources prepsource script on the remote host first step is therefore to write the new Fortran 90 sources Metcalf and Reed 1993 with an adaptation of the Doctor norm to the Fortran 90 described in Asencio et al 1994 We refer to the Meso NH web site in order to see how Meso NH routines are written In the next section we explain how to use the Meso NH source file manager to store these files in a form which allows its manipulation by the prepsource procedure section 7 3 Then we show how to compile these sources section 7 4 and how to use these binary objects in a simulation section 7 5 7 2 Meso NH source file management The Meso NH sources are managed using different elementary operations of the RCS source manager Essential RCS maintains a source file by creating new versions with a different version number 1 1 is the first version then 1 2 1 3 etc RCS also allows to extract modify or delete old versions of the code specific tools developped for the Meso NH system have basically the same functionalities of their RCS counterparts But they also permit the http mesonh aero obs mip fr mesonh 197 198 CHAPTER 7 MODIFY THE FORTRAN SOURCES identification of versions by character strings instead of the version identification number VID as in RCS This source file manager will help you to protect its developments against an unwa
249. rsion name are given in step 5 just in the way they would be given on the nh takeout command line We stress that mesonh does not perform any RCS manipulation Thus the modified file is purely local and not reintroduced into its home RCS library by prepsource Moreover it will be deleted from the local directory at the end of prepsource From step 7 onwards the sequence is basically the same than the one described above except for minor changes In step 8 the first default is now set up to the directory name in which prepsource has found the first source file In step 10 prepsource recognizes that the source file init f90 already exists on SIMUL it thus asks whether the file has to be overwritten this means reextracted from RCS or not In step 11 the user stops the general prepsource loop He finally enters the compilation options for the remote machine Step 13 consists of some information kindly printed by mesonh It indicates what source files are prepared for remote compilation here finally two source files which contained multi task code and which were customed for NSOURCE 1 and it says that a file inprepsource new was created which can be used for a future SILENTINTERACTIVE session with prepsource EXAMPLE OF THE PREPMODEL SCRIPT 225 Be careful when handling modules If the user tries to change modules that exist in the master library prepsource will exit Indeed modifications in the MODD and MODN modules from th
250. rst set of forcing field given by the user is lower or equal to the current date and time of the model run The forcing action of the last forcing field is remanant this is a way to impose a stationnary forcing When the current date and time of the model run is bounded by two successive forcing fields a simple linear interpolation in time is made Note that an available Newtonian relaxation forcing type on v and or 0 is exclusive from the other physical forcings forcing informations and soundings have to be added at the end of the free format part already written for CIDEAL CSTN or RSOU First the type of forcing and the number of time dependent forcing are given e keyword forcing type character 4 ZFRC means that the altitude of the forcing data are in height scale meters PFRC means that the altitude of the forcing data are in pressure scale Pascal e number of time dependent forcing integer The 1D forcing data are different from the one used to initialize the model because specific data have to be entered The data used to define each forcing are given sequentially in the following order one item per line e date and time of the forcing in the format year integer month integer day integer and time of the day real s e ground height real m e ground pressure real Pa ATTENTION in the MASDEV3 1 version either the ground height or the surface pressure was rea
251. s the order of namelists is free 1 Namelist NAM PGDFILE contains file names deti value CPGDFILE character LEN 28 PGDFILE YZOOMFILE character LEN 28 none YZOOMNBR character LEN 2 700 e CPGDFILE name of the input Physiographic Data File e YZOOMFILE optional name of the zoomned FM file output If the user does not specify this name or if YZOOMFILE CPGDFILE the code builds the zoomned FM file name as YZOOMFILE CPGDFILE zY ZOOMNBR e YZOOMNBR NumBeR which will be added to CPGDFILE to generate the name of the Zoomned FM file string of 2 characters 2 Namelist NAM MESONH DOM contains domain definition variables This was done during the PREP IDEAL CASE PREP REAL CASE step before the masdev4 6 version 5 4 ZOOM OF A PGD FILE ZOOM PGD 151 Fortran type default value integer input PGD domain integer input PGD domain integer NUNDEF if NUNDEF domain is in the middle of PGD integer NUNDEF if NUNDEF domain is in the middle of PGD e NIMAX number of grid points in I direction according to input file grid recovered by the new domain It must only be factor of 2 3 or 5 e NJMAX number of grid points in J direction according to input file grid recovered by the new domain It must only be factor of 2 3 or 5 e NXOR first point I index left to and out of the new physical domain e NYOR first point J index under and out of the new physical
252. storage machine for ectrans the directory has to exist e CLASS archive of extraction MARS parameter OD Operational Data E4 ERA4O reanalysis from 19570901 to 20020831 ER 15 reanalysis from 19781201 to 19940228 something else for example RD for data from Research Department In this case the parameters EXPVER and DATEGAP are also read e TYPE type of extraction for altitude and surface fields XX XX with XX MARS pa rameter in AN non initialized analysis IA initialized analysis only available for CLASS OD until 19960201 information can be found at http www ecmwf int research era 3information can be found at http www ecmwf int research era ERA 15 see the list at http www ecmwf int services archive d catalog 5 5 EXTRACTION OF ECMWF FILES EXTRACTECMWF 153 FG first guess short forecast from previous initialized analysis to smooth the fields FC forecast AN AN means AN for both altitude and surface fields IA FG means IA for altitude fields and FG for surface ones Orography and land sea mask surface fields are always requested with TYPE AN e DATE date written as YY YYMMDD e TIME validity time HH for analysis and first guess base time for forecast e STEP variable used for first guess or forecast files it corresponds to the length of forecast of the first guess file at verifying time i e forecast began at TIME STEP For time in 00 06 12 or 18 STEP must be 06 for time in 03 0
253. t has not an universal formula It is computed here for low level clouds according to Kunkel 1984 and the model COBEL 3 9 VISI Bac 144 7 Fire 0 88 6 17 Height and index of boundary layer top HBLTOP KBLTOP The boundary layer top is found by checking 00 02 and comparing it to the gradient between 5000m of height and the ground It is the same algorithm than in PREP REAL CASE used to shift variables when changing vertical grid which has been found to work fairly well from polar to saharian area More details can be found in the routine free atm profile f90 itself 6 3 2 Comparison with satellite observations comparison between model outputs and satellite observations provides an assessment of how well the model can reproduce the meteorological situation The model to satellite approach compares directly the satellite brightness temperatures BTs to the BTs computed from the predicted model fields Morcrette 1991 It has been first applied to Meso NH outputs for 6 3 VARIABLES AVAILABLE IN THE OUTPUT DIACHRONIC FILE 191 comparison with Meteosat observations in the infrared using a narrow band code Chaboureau et al 2000 Since the Masdev4_7 version the Radiative Transfer for Tiros Operational Ver tical Sounder RT TOV code version 8 7 Saunders et al 2005 is also available allowing the calculation of BT for a large number of satellites Computing satellite images from a Meso NH run You need an
254. t all layers the number of layers is the number of levels 1 In this case the level number can even be equal to 1 because the profile informations are linearly interpolated on the model grid without orography wind components 0 and humidity before the application of the Laplace relation to deduce the pressure and the vapor mixing ratio Thus the layers thicknesses are never too large to invalidate the Laplace relation CSTN Example of free part of PRE IDEA 1 nam 2006 06 06 21600 5 287 5 100240 200 1000 1500 3000 4000 10 20 25 30 35 2 10 12 5 11 5 15 80 84 85 79 87 0 01 0 014 0 015 0 016 3 3 4 The forced version keyword ZFRC or PFRC 3 8 FREE FORMAT PART A1 For idealized simulations a forced mode can be useful to impose the effects of a simplified large scale environment to the model solution This functionality works LFORCING TRUE in module MODD CONF when CIDEAL RSOU or CSTN see 5 2 10 and 5 3 and only in the case LCARTESIAN TRUE and LGEOSBAL FALSE for inclusion of a geostrophic wind forcing All forcing fields issue from spatial interpolation of chronological series of 1D data provided by the user onto the model grid They are prepared during the prep ideal case sequence and are stored in the LFI files for further use in case of RESTART model run The forcing fields can be time dependent Application of the forcing begins as soon as the date and time of the fi
255. t of total liquid temperature flux terms of the equation of 2 lt wo gt lt w 0l gt gt are computed and stored in the diachronic group BU WTHL comments made for the total Tke equation are valid here ADVM ADVR 7 5s 10 5 lt gt lt Ua gt lt 00 gt u sz wuw Ot Oxo l 06 Ow ee es ae lt 00 gt Org BU P TP DPR ADV PRES m m 1 Op lt l gt lt gt lt 46 gt tos Qua lt gt EN lt 8010 gt ADD gt TR 12 lt 2 gt 2 lt a gt DP dis lt 0 gt lt Uaw gt lt 2 gt o lt w gt e n lt Wy UG gt lt Urs usu gt a SBGT field notation in processus dim comments diac file name BU WTHL SBGDPM dyn prod by mean gradient w 0 gt BU WTHL SBG DPR 7 dyn prod by resolved fluctuations O 4 2 0 Soe WO gt curl lt gt Op SU d BU WTHL SBG TR subgrid turbulent transport U_WTHL SBG PRES z t subgrid pressure correlation term lt gt BUAWTHL SEG TP EST budget of lt gt 44 BUDGET OF RESOLVED SUBGRID TURBULENT QUANTITIES 115 field notation in processus dim comments i ie me e MEE of lt 00 gt p
256. te on line dust emissions see surface namelists Radiative direct effects are automatically deduced from an interpolation table of SHDOM ra diative code Mie Fortran type default value LDUST logical LVARSIG logical LSEDIMDUST logical NMODE DST integer e LDUST switch to activate passive dust aerosol e LVARSIG switch to activate variable standard deviation for each dust modes e LSEDIMDUST switch to activate dust sedimentation e NMODE DST number of lognormal dust modes maximum of 3 modes 56 CHAPTER 4 PERFORM A MESONH SIMULATION 4 2 4 Namelist NAM DYN global parameters for the dynamics Fortran type default value XSEGLEN real XASSELIN real XASSELIN SV real LCORIO logical LNUMDIFF logical LZDIFFU logical XALKTOP real XALZBOT real It contains the dynamics parameters common to all models They are included in the module MODD DYN XSEGLEN Segment Length in seconds corresponding to the duration of the segment simulation XASSELIN Amplitude of the Asselin temporal filter for meteorological variables XASSELIN SV Amplitude of the Asselin temporal filter for scalar variables LCORIO Switch to set the Coriolis parameters f and f to zero the Earth rotation is taken into account FALSE the Earth rotation effects are neglected LNUMDIFF Switch to activate the numerical diffusion XT4DIFF in NAM_DYNn defines the intensity of this diffusion LZDIFFU Switc
257. terpolated fields of CINIFILE This allows to keep finest information when defining a new finest domain to follow atmospheric system 3 NAM BLANK see Perform a MESONH simulation for description 4 Namelist NAM SPAWN SURF dett value LSPAWN SURF TRUE e LSPAWN SURF flag to choose to or not to perform the interpolation of all the sur face fields both physiographic and prognostic fields Note that these interpolations are performed by externalized surface facilities However no specific namelist is required for this operation LSPAWN SURF TRUE the interpolations are made LSPAWN SURF FALSE the interpolations are not made and therefore no surface field will be present in the output spawned file 172 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES 5 9 If you want surface fields only PREP SURFEX PREP_SURFEX performs the interpolations of surface fields from one grid to another What s going in and out e Input the prepmodelrc file a file containing the surface 2D variable fields hereafter called input file it can be either an GRIB file obtained from extractecmwf or extractarpege a MESO NH file obtained with SPAWNING for example a physiographic data file it can also be a complete MESO NH file the file PRE REALI nam which contains the directives for PREP SURFEX e Output the FM file containing the physigraphic and pronostic surface fields 5 9 1 The prepm
258. test on 0 profile XLES_TEMP_SAMPLING time seconds between two samplings of the LES profiles and non local quantities XLES TEMP MEAN START time seconds from the begining of the simulation at which the averaging begins If not defined no averaging is performed 4 2 THE INPUT EXSEG N NAM FILE 97 XLES_TEMP_MEAN_END time seconds from the begining of the simulation at which the averaging ends If not defined no averaging is performed XLES TEMP MEAN STEP time step seconds for averaging LLES_CART_MASK logical switch to compute the LES diagnostics only inside a cartesian subdomain Both local and non local quantities can be computed NLES IINF lower i index of the cartesian subdomain Default is physical domain left boundary NLES ISUP upper i index of the cartesian subdomain Default is physical domain right boundary NLES JINF lower j index of the cartesian subdomain Default is physical domain bottom boundary NLES_JSUP upper index of the cartesian subdomain Default is physical domain top boundary LLES NEB MASK logical switch to compute the LES diagnostics separately inside and outside the model columns where clouds are present Only local quantities can be com puted LLES CORE MASK logical switch to compute the LES diagnostics separately inside and outside the model columns where cloud core is present Only local quantities can be computed LLES MY MASK logical switch to compute
259. the LES diagnostics on a mask defined by the user The mask is 2D horizontal It must be coded at the beginning of the LES monitor routine Only local quantities can be computed on this mask 98 CHAPTER 4 PERFORM A MESONH SIMULATION 4 2 44 LES budgets Namelist NAM PDF Fortran type default value LLES PDF logical NPDF integer XTH PDF MIN real XTH_PDF_MAX real XW_PDF_MIN real XW_PDF_MAX real XTHV_PDF_MIN real XTHV_PDF_MAX real XRV_PDF_MIN real XRV_PDF_MAX real XRC_PDF_MIN real XRC PDF MAX real XRR_PDF_MIN real XRR PDF MAX real XRI_PDF_MIN real XRILPDF_MAX real XRS_PDF_MIN real XRS_PDF_MAX real XRG_PDF_MIN real XRG PDF MAX real XRT_PDF_MIN real XRT_PDF_MAX real XTHL PDF MIN real XTHL_PDF_MAX real Each PDF includes NPDF intervals number between X_PDF_MIN and X_PDF_MAX e LLES_PDF Logical switch for pdf computation e NPDF Number of PDF intervals e XTH PDF MIN Minimum of potential temperature pdf XTH PDF MAX Maximum of potential temperature pdf e XW PDF MIN Minimum of vertical velocity pdf e XW PDF MAX Maximum of vertical velocity pdf XTHV_PDF_MIN Minimum of virtual potential temperature pdf XTHV PDF MAX Maximum of virtual potential temperature pdf XRV PDF MIN Minimum of vapor mixing ratio pdf XRV PDF MAX Maximum of vapor mixing ratio pdf 4 3 LES DIAGNOSTICS 99 e XRC PDF MIN Minimum of cloud mixing ratio pdf e XRC_PDF_MAX Maximum of cloud mixing ratio pdf e XRR
260. the model on a smaller domain than the one of the PGD file defined at steps 1 and 2 for the model 3 you must use this program SPAWNING when you want to start the model 3 you must use this program to compute the horizontal interpolations from the model 2 to the model 3 It is used only once for the initialisation of model 3 142 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES 12 PREP REAL CASE It is used only once to compute the initial file for the model 3 Do not change the vertical grid 13 MESONH here is now your complete nested run step g 3 o 4 5 9 13 s 4 1 6 7 8 9 13 0 10 11 12 3 Figure 5 5 Exemple of a grid nesting simulation with 3 nested models e The following initialisation and gridnesting sequence is shown here for three models model 2 included in model 1 and model 3 included in model 1 Model 3 has the same resolution as model 2 and is started after model 2 to follow atmospheric system figure 5 6 1 PREP_PGD one physiographic data file for the model 1 definition of projection resolution domain one physiographic data file for the models 2 and 3 same projection definition of resolution domain 2 PREP NEST PGD this program checks all the two PGD files at the same time and imposes the conformity between them 3 extractecmwf or extractarpege it extracts the surface and altitude fields for one date for mo
261. the model under the form of a set of Fortran 90 namelists We choose an academic case i e simplified orography and atmospheric conditions for the generation of initial conditions because it makes easier the presentation because less models and file types are involved The real case presentation is delayed to Chapter 5 In Chapter 4 the degrees of freedom of the model itself are presented We also recall where informations are read either to start the simulation FM file or to pilot the MESONH model from a Larger Scale Model In Chapter 5 we present in the same way the real case study and its own degrees of freedom Several programs are involved to generate the initial MESONH FM file used either in a mono model simulation or in a multi model simulation with grid nesting technics The Chapter 6 lists the diagnostic variables which can be obtained after a simulation In Chapter 7 the file manager is presented and we show how to modify or add Fortran sources to the MESONH sources We also describe how to merge these new elements with the Meso NH sources In Annexe 1 the FM file structure is described in details 10 CHAPTER 1 INTRODUCTION Chapter 2 The MESONH procedures except prepsource 2 1 Presentation 2 1 1 What can you do with the MESONH procedures MESONH procedures provide a complete UNIX shell script which allows the MESONH user to compile a subroutine to prepare a numerical simulation to run the mod
262. the remote host but this step is very machine dependant local host and remote host Anyway limits for the duration values of memory processors and nodes of the job are typical informations which must be provided to correctly configurate this step 20 CHAPTER 2 THE MESONH PROCEDURES EXCEPT PREPSOURCE Chapter 3 Prepare an initial MESONH file for an ideal case 3 1 Overview of PREP IDEAL CASE functionalities The PREP_IDEAL_CASE program prepares a MESONH file that contains all the parameters and fields necessary for the execution of the MESONH model Specifically the grid parameters the initial fields and the geophysical fields are included in this file It is possible using this program to generate idealized fields defined by few parameters This is useful for e debugging purpose e simulation starting from an idealized initial field The generated initial conditions are produced analytically leading to quasi 1D fields or 3D fields or a single profile build with either e layers of constant Brunt Vaisala frequency shear and humidity or e a Radiosounding and ideal surface fields e a Radiosounding and real physiographic fields e a Radiosounding and real and ideal surface fields at the same time For these latter cases the initial fields may be hydrostatically or geostrophically balanced or not For these fields to satisfy the anelastic constraint a final correction is applied to them The interaction
263. tialization or coupling file fine grid fine resolution new vertical grid Sp surface pressure spec zs spectral orography large scale orography Ts temperature of surface stands for all other surface variables temperatures water and snow contents PGD fields stands for cover fractions orographic vegetation and soil physiographic fields Figure 5 1 schematic view of the interactions between the different files during the initialization sequence of a real case simulation from a global file of ECMWF or METEO FRANCE archive 5 1 OVERVIEW OF THE INITIALIZATION SEQUENCES 137 data files NES LUE prepmodelrc PRE_PGD1 nam large domain Creation of PhysioGrafic Data file prepmodel PREP_PGD PhysioGrafic Data file extraction from extraction from ECMWF Meteo France extractecmwf extractarpege 1 GRIB file 1 Horizontal and PRE REALI nam vertical MESO NH domain interpolations prepmodel prepmodelrc PREP REAL CASE MESO NH file Figure 5 2 initialization sequence of a real case simulation from a global file of ECMWF or METEO FRANCE archive 138 CHAPTER 5 INITIALIZATION OF MESO NH FOR REAL CASES 5 1 2 Initialization for a one model run from a MESO NH file In this case the initial and coupling files are interpolated from a previous MESONH run at coarser resolution It is again necessary to produce separately one initial fil
264. ticle de donnees STORAGE TYPE 24 mots position 1637 a 1660 4 eme article de donnees IMAX 23 mots position 1661 a 1683 5 eme article de donnees JMAX n 23 mots position 1684 a 1706 6 eme article de donnees KMAX 23 mots position 1707 a 1729 7 eme article de donnees THINSHELL 23 mots position 1730 a 1752 8 eme article de donnees LATO 23 mots position 1753 a 1775 9 eme article de donnees LONO 23 mots position 1776 a 1798 10 eme article de donnees BETA nm 23 mots position 1799 a 1821 11 article de donnees XHAT ve 184 mots position 1822 a 2005 12 eme article de donnees YHAT n 25 mots position 2006 a 2030 13 eme article de donnees ZHAT 68 mots position 2031 a 2098 14 eme article de donnees ZS 508 mots position 2099 a 2606 15 article de donnees DTCURATDATE 25 mots position 2607 a 2631 16 eme article de donnees DTCURATIME 23 mots position 2632 a 2654 17 eme article de donnees DTEXPATDATE n s 25 mots position 2655 a 2679 18 eme article de donnees DTEXPATIME n 23 mots position 2680 a 2702 19 eme article de donnees DTMOD TDATE a 25 mots position 2703 a 2727 20 article de donnees DTMODATIME M 23 mots position 2728 a 2750 21 article de donnees DTSEGATDATE 25 mots position 2751 a 2775 22 article de donnees DTSEG TIME an 23 mots position 2776 a 2798 23 eme article de donnees CARTESIAN n 23 mots position 2799 a 2821 24 eme article de d
265. tion 398770 a 413695 45 eme article de donnees LSXRRM id 298 mots position 413696 a 413993 46 eme article de donnees LSYRRM 14926 mots position 413994 a 428919 47 eme article de donnees LSXRIM 298 mots position 428920 a 429217 48 eme article de donnees LSYRIM 14926 mots position 429218 a 444143 49 eme article de donnees LSXRSM 298 mots position 444144 a 444441 50 article de donnees LSYRSM 14926 mots position 444442 a 459367 51 article de donnees LSXRGM we 298 mots position 459368 a 459665 52 eme article de donnees LSYRGM 14926 mots position 459666 a 474591 53 eme article de donnees UT 22378 mots position 474592 a 496969 54 eme article de donnees VT 22378 mots position 496970 a 519347 55 eme article de donnees WT 22378 mots position 519348 a 541725 8 2 THE MESO NH FILES 56 57 eme 58 eme 59 eme 60 eme 61 eme 62 eme 63 eme 64 eme 65 eme 66 eme 67 eme 68 eme 69 eme 70 eme Ti eme 72 eme 73 eme 74 eme 75 eme 76 eme 77 eme 78 eme 79 eme 80 eme 81 eme article de donnees 22378 mots position 541726 a 564103 article de donnees TKET 22378 mots position 564104 a 586481 article de donnees RVT 22378 mots position 608860 a 631237 article de donnees RCT 22378 mots position 631238 a 653615 article de donnees RRT 22378 mots position 653616 a 675993 article de donnees RIT 22378 mots position 675994 a 698371 arti
266. tions for the convective scheme Of model quus cud 70 4 2 19 Namelist NAM PARAM_RADn options for the radiations of model n 71 4 2 20 Namelist NAM TURBn turbulence parameters for model n 74 4 2 31 SURFACE SCHEMES namelists of the externalized surface T5 4 2 22 CHEMISTRY scheme Namelist NAM CH MNHCn control of MNHC 76 4 2 23 Namelist NAM CIL ORILAM 78 4 2 24 Namelist NAM CH SOLVERn control stiff solvers for modeln 81 4 2 25 Simulation of the fly of balloons or aircraft in the model fields 81 22 20 Profilers and stations s s Sr SR qe RR uer E dur 82 4 2 27 Namelist NAM SERIES temporal series in diagnostic file 82 4 2 28 Namelist NAM SERIESn temporal series in diagnoctic file of model n 83 4 2 29 Namelist NAM BUDGET budget box 84 4 2 30 Namelist NAM BU RU budget for 86 4 2 31 Namelist NAM BU RV budget 87 4 2 32 Namelist NAM BU RW budget for W 87 4 2 33 Namelist NAM BU RTH budget for 88 4 2 34 Namelist NAM BU budget for 89 4 2 35 Namelist NAM BU RRV budget for 89 4 2 36 Namelist NAM BU RRC budget for cloud water 90 4 2 37 Namelist NAM BU RRR budget for rain water 91 4 2
267. type of grid YZGRID_TYPE is set to MANUAL in the namelist and only the number of levels NKMAX is also used in it The variables of this namelist are LTHINSHELL logical FALSE NKMAX integer 10 YZGRID TYPE 6 characters PUNCTN ZDZGRD real 300 ZDZTOP real ZZMAX_STRGRD real ZSTRGRD real ZSTRTOP real LSLEVE logical XLENI real XLEN2 real e LTHINSHELL switch for the thinshell approximation logical e NKMAX number of points in z direction of the required physical domain The total size of the array written in initial file will be NK MAX 2JPVEXT JPVEXT is fixed to 1 for the present version of Meso NH e YZGRID_TYPE type of vertical grid definition FUNCTN the vertical grid is given by a regular logarithmic function whose varia tion is determined by the values of free parameters ZDZGRD ZDZTOP ZSTRGRD ZSTRTOP ZZMAX_STRGRD described below MANUAL the levels are explicitly given in the free formatted part by entering the heights of the different levels from the 2 to KMAX 2 Therefore you only enter KMAX 1 values because the level under the ground i e K 1 is at the same distance from the ground K 2 as the first level above the ground K 3 Note also that the K KMAX 2 level represents the model top In this case the free parameters ZDZGRD ZDZTOP ZSTRGRD are not used e ZDZGRD mesh length in z direction near the ground e ZDZTOP mesh length in z direct
268. value e LBU_JCP Logical switch to average or not in the J direction all the budget terms for CBUTYPE CART e LBU ICP Logical switch to average or not in the I direction all the budget terms for CBUTYPE CART The description of the budgets for every prognostic variable is given below Because all the budgets are performed in the same way we give here some details on the way to select or cumulate the different source terms Firstly there is a switch to activate or not the budget of a given prognostic variable It should be noted that the budget terms for the variable V have the dimension of 0 pv Ot Then all the source terms computed in the model for this prognostic variable can be selected according to the following rules 86 CHAPTER 4 PERFORM A MESONH SIMULATION e NSOURCE_TERM 0 if you do not want to take into account this source term it happens when you are only interested in some terms of a budget and not to the whole budget and you discard the others e NSOURCE TERM 1 if you select this source term and it is the first element of a set of source terms which are cumulated e NSOURCE TERM 2 if you select this source term but it is not the first element of a set of source terms which are cumulated Note that the source terms which come just after the last of this set of source term must be discarded NSOURCE TERM 0 or be the first element of a new cumul NSOURCE TERM 1 We now give every namel
269. value of standard deviation for mode J accumulation mode 6 NAM BLANK see Perform a MESONH simulation for description 7 Namelists of the externalized surface for PREP REAL CASE surface initial fields are produced by externalized surface facilities So you are invited to refer to the documentation of the surface For PREP REAL CASE you must fill the following namelists NAM PREP SURF ATM NAM PREP SEAFLUX if you chose to use the SEAFLX scheme NAM PREP WATFLUX if you chose to use the WATFLX scheme NAM PREP TEB if you chose to use the TEB urban scheme NAM PREP ISBA if you chose to use the ISBA scheme CAUTION a b Note that all namelists can be void but only if the initial file name for HATMFILE you provide in namelist FILE NAMES contains externalized surface fields If the file HATMFILE does not contains externalized surface fields you must fill at least namelist NAM PREP SURF ATM if you want to initialize the surface prog nostic fields from a input file You can also define more precisely the surface fields by using the namelists for each scheme 5 7 VERTICAL INTERPOLATIONS PREP REAL CASE 167 8 Free formatted part Vertical grid This part is optional in the file read only if YZGRID_TYPE MANUAL It must begin by the keyword ZHAT In this case NKMAX 1 levels are written in meters in free format after the keyword from ground level generally 0 to rigid top l
270. vection scheme used for the tracer variables The following op tions are possible see the Scientific Documentation for more details CEN2ND 2nd order advection scheme CENtred on space and time It does NOT guarantee the sign preservation CENATH 4th order advection scheme CENtred on space and time It does NOT guarantee the sign preservation FCT2ND 2nd order advection scheme CENtred on space and time It is POSITIVE definite MPDATA 2nd order advection scheme uncentred on space and time It is POSITIVE definite PPM_00 PPM advection scheme without constraint IPPM 01 Monotonic version of PPM It is POSITIVE definite e NLITER number of iterations that the MPDATA is applied NLITER 1 donor cell or upstream scheme 64 CHAPTER 4 PERFORM A MESONH SIMULATION 4 2 12 Namelist NAM CONFn configration of model n Fortran type default value LUSERV logical TRUE LUSECI logical FALSE NSV_USER integer 0 It contains the model configuration parameters specific for the model n They are included the module MODD CONFn e LUSERV Switch to use vapor water prognostic variable r TRUE rv is present FALSE rv is not allocated e LUSECI Switch to use Pristine Ice diagnostic variable C TRUE C is present FALSE C is not allocated e NSV USER Number of user passive scalar variables Caution Scalar variables needed for the 2 mo

Download Pdf Manuals

Related Search

Related Contents

  Philips 170T4FS  Use and Care Guide  Manual de instruções Extensão do cabo HVC      

Copyright © All rights reserved. Failed to retrieve file