USER GUIDE FOR HARMONIE MUSC
Contents
1. ALADIN echo ALADIN job running echo set x export DR_HOOK 1 MASTER c001 vmeteo maladin e exp t TSTEP f NSTOP a ADVEC gt lola 2 gt amp 1 rm fort 4 Is 1 set x echo echo Listing d execution pour partie non parallelisee fichier lola echo Listing for the not parallelised part file lola echo set x cat lola if a NODE 001_01 then for file in NODE do set x echo echo Listing d execution pour partie parallelisee fichier file echo Listing for the parallelised part file file echo set x cat file done fi dl HH kkk k kok kok kok k k k Ro k kok deae kok kop kok K K qe Envoi des resultats du modele Save model results FH BRAG k k AR k k k k do kok de o oco k K kok k k kok ooi eoe set x echo echo Sauvegarde des fichiers historiques echo Historic files saving echo set x define name of output files on delage mkdir FILE_PATH OUTPUT OUTNAME FILE_PATH OUTPUT OUTNAME save model results mv Out FILE_PATH OUTPUT OUTNAME mv OUTPUT FILE_PATH OUTPUT OUTNAME mv NODE FILE_PATH OUTPUT OUTNAME set x echo echo Fichiers presents sur le workdir WRK_PATH echo Present files on the workdir WRK_PATH echo set x ls 1 LELLLLLLLLLLLLLLLLLLLLLLLELLLLLLLLLI Nettoyage final Final Cleaning ee Xk K K2. k K KK K K ok oK K 2K K 3K ke K K K K ke Xk K 3K ok K KK ee oK K set x echo echo Nettoyage final du workdir WRK_PATH echo Final cleaning of the workdir WRK_PATH echo set x rm namprovi rm MASTER suppression des fichiers ALADIN removal of the ALADIN files for file in ICMSH DHF do rm file done for file in AROMOUT do rm file done for file in ELSCF do rm file done rm ECHIS rm ifs disp rm ifs stat rm lola if a NODE 001_01 then for file in NODE do rm file done fi rm f core cd rmdir WRK_PATH cd OUT_PATH In s basemusc Ifa2ascii sh sh lfa2ascii sh SOUTNAME 3 3 Testinput and expected output The test data for your experiment are in the package musctest tgz Please open it in your own home directory cd cp home hirlam harmonie MUSC musctest tgz tar xvfz musctest tgz and copy also the reference scripts cd musc cp r home hirlam harmonie MUSC scr cp scr mur HOME bin Please check the scripts mur and Runmusc for the directory definitions and usage alternatives Now you are ready to say scr mur MUTEST The script mur will set your experiment environment and bring you to the new experiment directory HOME musc EXP MUTEST It is also called MU WD for your convenience With the test package you already got test data to your new MU WD mur creates a defitinion file tool path and advises you to run it as tool pa
3. E makeup e A HARMONIE file handling utility gl may be applied for extraction of input data from the three dimensional HARMONIE files 2 Setup of harmonie MUSC 2 1 Structure of the code harmonie MUSC branch contains HARMONIE cycle37h source code dedicated utilities and scripts The idea is that all basic HARMONIE source code should be available together with some MUSC specific additional tools for handling of input output and MUSC experiments It is suggested that in the workstation environment a structure with reference directory possibly under a reference user s home directory open for reading to all users and the user home and scratch directories are kept separetely This is the structure used in HIRLAM and HARMONIE 3D experiments with or without mini SMS user interface For example home hirlam harmonie_MUSC src scr reflib util bin home rontu musc scr general scripts EXP EXPm EXPn MU_WD src modified source codes OUTPUT output directories scratch EXPm EXPn MU_WRK lib src compiled and source code bin compiled MASTER 2 2 Getting the source code utilities and test data cd home hirlam svn co https svn hirlam org branches harmonie_MUSC You need a password for hirlam org for this With subversion checkout you will get the a README file source code utilities scripts and a testdata package 2 3 Compilation with makeup First please check you have needed software in your computer You
4. MIDITY R4 1 PTS SURFACE TEMPERATURE R4 1 PQS SURFACE SPECIFIC HUMIDITY R4 1 PFRSODS SURFACE DOWNWELLING SW R4 1 PFRTHDS SURFACE DOWNWELLING LW R4 1 PITM R4 nlev 1 PSTRTU R4 nlev 1 PSTRTV R4 2 PFRSOC SW CLEAR SKY NET RADIATION TOA SFC R4 2 PFRTHC LW CLEAR SKY NET RADIATION TOA SFC R4 1 THETAS SURFACE POTENTIAL TEMPERATURE R4 1 HCLA hauteur de la couche limite atmospherique R4 1 HCLA_AY1996 R4 nlev 1 ZCP_flux R4 nlev 1 ZLH flux LATENT HEAT R4 1 ZUSTAR FRICTION VELOCITY R4 1 SAT_DEF_0 15 R4 1 SAT_DEF_1 15 R4 1 SAT_DEF_2 5 4 5 R4 1 SAT_DEF_1 4 3 R4 1 SAT_DEF_QV_0 15 R4 1 SAT_DEF_QV_1 15 R4 1 SAT_DEF_QV_2 5 4 5 R4 1 SAT_DEF_QV_1 4 3 R4 1 ZCLDTOP R4 1 ZCLDBAS R4 1 ZMAXFRAC
5. R4 nlev PR TYPE LEVELS NAME DESCRIPTION R4 nlev PVERVEL VERTICAL VELOCITY R4 nlev ZWA R4 nlev ZOMEGA LARGE SCALE VERTICAL VELOCITY R4 nlev 1 PEMTD TOTAL DOWNWARD LONGWAVE EMISSIVITY R4 nlev 1 PEMTU TOTAL UPWARD LONGWAVE EMISSIVITY R4 nlev 1 PTRSO TOTAL SHORTWAVE TRANSMISSIVITY R4 1 PCLON COSINE OF LONGITUDE R4 1 PSLON SINE OF LATITUDE R4 nlev 1 PFRSO NET SW RADIATIVE FLUX R4 nlev 1 PFRTH NET LW RADIATIVE FLUX R4 1 PTCLS 2M TEMPERATURE R4 1 PUCLS 10M U COMPONENT R4 1 PVCLS 10m V COMPONENT R4 1 PQCLS 2M SPECIFIC HUMIDITY R4 1 PCLCT TOTAL CLOUD COVER R4 1 PCLCH HIGH CLOUD COVER R4 1 PCLCM MEDIUM CLOUD COVER R4 1 PCLCL LOW CLOUD COVER R4 1 PCLCC CONVECTIVE CLOUD COVER R4 1 VENTCLS 10M WIND SPEED R4 1 DIRCLS 10M WIND DIRECTION R4 1 PFCLL LATENT HEAT FLUX OVER LIQUID WATER OR WET SOIL R4 1 PFCLN LATENT HEAT FLUX OVER SNOW OR ICE R4 1 PFCS SENSIBLE HEAT FLUX AT SURFACE LEVEL R4 nlev PNEB FRACTIONAL CLOUDINESS FOR RADIATION TYPE LEVELS NAME DESCRIPTION R4 1 ECT_INT R4 1 LWP LIQUID WATER PATH R4 1 IWP ICE WATER PATH R4 1 CWP TOTAL LIQ ICE PATH R4 1 WVP WATER VAPOR PATH R4 nlev 1 PFPLSL LIQUID PRECIPITATION R4 nlev 1 PFPLSN SNOW PRECIPITATION R4 nlev 1 PFPLSG GRAUPEL PRECIPITATION R4 nlev 1 PFPLSHL HAIL PRECIPITATION R4 nlev 1 PREC_TOT TOTAL PRECIPITATION R4 nlev PQICE SPECIFIC ICE CONTENT R4 nlev PQLI SPECIFIC LIQUID WATER CONTENT R4 nlev PRH RELATIVE HU
6. T PATH test d SWRK PATH Il mkdir p WRK PATH cd WRK PATH if cycle 37t1 then test e namfil ll echo echo ERROR namelist missing exiting exit amp amp cp namfil namprovi 3t Fichier et namelist SURFEX cp SFILE PATH AROME PREPSURFEIfi TEST Ifi cp FILE_PATH EXSEG1 nam EXSEGI nam fi run time optimization options export ZOPT WIl d100 g250 e1 date ADVEC eul ADVEC sli cat lt namprovi Fichier initial test e inifil amp amp bin cp Sinifil ICMSHLOTSINIT ll echo echo ERROR inifile missing exiting FILENAME 1 ICMSHLO7SINIT FILENAME2 ICMSHLOTSINIT init file and coupling file bin cp FILENAME1 ICMSH expINIT d BESSA SS A k k k k k k k k kok si ois re sls sv sje k KK kok K K Execution oie ois X kokk E SEE SEE K K oft SE HE of ok K SE KK HE K SE ok ok K ok ok ok ok K K set x echo echo Acquisition de executable echo Get the executable file echo set x Get the executable file if SMYOWNEXE then In s MYOWNEXE MASTER chmod 755 MASTER else echo GIVE MYOWNEXE exit get SGOUROUBIN mv GOUROUBIN MASTER fi set x echo echo Copie de la namelist sur fort 4 echo Namelist copy on fort 4 echo set x cat namprovi awk print 1 FS gt fort 4 set x echo echo Execution du job
7. USER GUIDE FOR HARMONIE MUSC This draft is provided for the MUSC working week 29 11 2 12 2011 as a template for further development Ulf Andrae Eric Bazile Cisco de Bruijn Laura Rontu Sami Saarinen E mail Laura Rontu fini fi Abstract aaaa 1 INTRODUCTION Single column models are used for development and testing of physical parametrizations of NWP models In princi ple such models contain in principle the whole source code of the prognostic model This should allow to run idealised experiments focusing on the atmospheric physics in a simplified framework but based on the same code that is used for weather prediction and research Simplified framework means e Time integration is done in a single atmospheric column e The initial state possible atmospheric forcing during time stepping and surface description are provided as input e Horizontal advection is not included and the model dynamical formulations are applied in minimal extent Because of the simplifying assumptions a single column model is not suitable for real forecasting Its value is in the possibility of study sensitivity of the model to different formulations of the physical parametrizations in different realistic atmospheric conditions It is possible to quickly run experiments with any combination of physical parametrizations in a workstation environment Application of very high resolution in vertical should be possible On the other hand any results and suggestion
8. le plotting grads matplotlib metview for data analysis and plotting TYPE LEVELS NAME DESCRIPTION C 1 INDICE EXPERIENCE R4 nlev RHO DENSITY R4 nlev 1 RHO flux DENSITY FLUX I4 1 KLEV NO VERTICAL LEVELS R4 1 TSPHY PHYSICS TIMESTEP I4 1 NINDAT DTG AS INTEGER I4 1 NSSSSSa initial time in seconds e g for 12h 43200 I4 1 KCLPH LEVEL OF PBL R4 1 RSTATI NUMBER OF SECONDS SINCE START OF THE MODEL R4 nlev PU U COMPONENT R4 nlev PV V COMPONENT R4 nlev PVENT WINDSPEED R4 nlev PDIRVENT WINDDIRECTION R4 nlev PT TEMPERATURE R4 nlev PQ SPECIFIC HUMIDITY R4 nlev ZH HEIGHT OF HALF LEVELS R4 nlev THETA POTENTIAL TEMPERATURE R4 nlev THETAV VIRTUAL POTENTIAL TEMPERATURE R4 nlev THETAL_KE R4 nlev THETAL_BS R4 nlev THETAL LIQUID WATER POTENTIAL TEMPERATURE R4 nlev THETAVL LIQUID WATER POTENTIAL VIRTUAL TEMPERATURE R4 nlev PECT R4 nlev PQI SPECIFIC HUMIDITY ICE R4 nlev PQL SPECIFIC HUMIDITY LIQUID WATER R4 nlev PQSN R4 nlev PQR R4 nlev PQG I4 1 KGLI I4 1 KGL2 I4 1 KSGST 14 1 KSTEP TIMESTEP I4 1 KCSS R4 1 PMUO SOLAR ZENITH ANGLE R4 1 PGEMU R4 1 PGELAM R4 nlev 1 PAPHI geopotential height gz at half levels R4 nlev 1 PAPRS hydrostatic pressure at half levels R4 nlev PAPHIF geopotential height gz at half levels R4 nlev PAPRSF hydrostatic pressure at full levels R4 nlev PALPH COEFFICIENTS OF THE HYDROSTATICS R4 nlev PDELP PRESSURE DIFFERENCE ACROSS LAYERS R4 nlev PLNPR R4 nlev PRDELP R4 nlev PCP
9. s based on single column studies need careful testing in the full model A special approach and additional tools are required for handling input and output of a single column model The first version of the Single Column Unified Model SCUM was derived from the full IFS code cy32 by Sylvie Malardel in 200 Later SCUM was renamed for aestethical reasons to MUSC Model Unifie Simple Colonne MUSC has been further developed and applied in Meteo France by Eric Bazile in KNMI by Cisco de Bruijn et al by Javier Calvo et al in AEMET Versions based on IFS cycle 33 and 35 have been defined A user guide was written and updated at KNMI During this development MUSC started to diverge from the three dimensional reference code The first of HARMONIE MUSC is being built on IFS cycle 37 taking a MUSC export version of cycle 37t1 from Meteo France as a starting point The aim is to bring MUSC back to the original idea of being fully compatible with the three dimensional HARMONIE Presently a separate branch harmonie_MUSC exists in hirlam org subversion repository In the future this branch should converge with the trunk of HARMONIE In the first step of conversion it should be possible to run any three dimensional HARMONIE experiment based on the code of the MUSC branch The specific features of the HARMONIE MUSC code include e Source code is maintained and made available in the hirlam org repository e Compilation of the code is based on the HARMONI
10. th to make utilities and scripts available in your PATH Do not forget to do as advised In MU WD you should say Runmusc gt musc log to run Eric s test case and use the ddh tools with Eric s script lfa2ascii sh to convert main output files to ASCII You can have a look at them or try something However we are working with the next step i e visualisation tools for the lot of ASCII files of the instant and accumulated profiles or different variables 4 Preparation of input for experiments You might want to produce your own atmospheric column for MUSC input instead of the initfile provided by the test package For this gl 1 n naminterp fa ICMSHHARM 0006_aladin you should have got naminterp fa and ICMSHHARM 0006_aladin with the test package newest version gl should produce you a file CMSHHARM 0006_aladin fa which is ready for input to your experiment Not that gl should be the one recently compiled in refmusc bin Please copy the reference Runmusc to Runmusc testinput and modify it by changing klev value as needed command gl 1 ICMSHHARM 0006_aladin grep lev gives the needed number chmod 755 Runmusc_testinput and Runmusc_testinput And yes copy the ICMSHHARM 0006_aladin fa to a file with the name Runmusc testinput expects Check the directory OUTPUT for new files Works Got the idea By the way it might be in the future better to create a directory MU_WRK muscin and go there for input preparation to avoid e
11. will need gcc and gfortran gt v 4 4 mpi libraries e g in Debian libmpich2 dev lapack blas library e g in Debian liblapack dev ksh flex bison Please ask your systems package installation tool to bring these to you e g in Debian or Ubuntu apt get install bison etc There are MUSC configurations for debian ubuntu redhat gfortran work with fedora has been started The configuration files are in util makeup Next build the code cd home hirlam harmonie MUSC util makeup build config debian gfortran This will produce the main executable file home hirlam harmonie MUSC bin MASTER which can be later linked to your experiment In addition several utilities will be compiled and added to the same directory You will need DDH toolbox Compile it separately cd home hirlam harmonie MUSC util ddhtoolbox tools sh install sh Please find documentation of toolbox usage in home hirlam harmonie MUSC util ddhtoolbox documentation 3 Running a MUSC experiment 3 1 Structure of an experiment Figure 1 Structure of a MUSC experiment from input to visualisation 3 2 Script Runmusc exp ARPE TSTEP 300 NSTOP h12 mod ARPEGE klev 70 ver 25 cycle 37tl expmusc basemusc EXP EXP NAM_PATH expmusc FILE_PATH expmusc namfil NAM_PATH namus_ cycle_ mod inifil FILE_PATH initfile_ ver_L klev OUTNAME mod_ ver_ TSTEPs_L klev_ exp OUT_PATH expmusc OUTPUT WRK_PATH wrkmusc rundir test d OUT PATH Il mkdir p SOU
12. xtra mess with the files 1 Available tools gl and acadfa pgd prep 4 4 2 Preparation of atmospheric column 4 3 Preparation of surface dat 5 Modifying of namelists for MUSC experiment 5 1 Run time namelist 5 2 SURFEX special definitions 5 3 Example for AROME ALARO ARPEGE 6 Modifying source code of an experiment 6 1 You can modify harmonie MUSC source codes in your MU WD For example to modify apl arome F90 in an experiment called MYOI you do this mkdir p src arp phys dmn cp Srefmusc src arp phys dmn apl arome F90 src arp phys dmn cd src arp phys dmn modify apl arome F90 as wanted save cd MU WD mur MYO1 recompile Your modifications will be copied over the reference code in MU_WRK lib src arp phys_dmn and compiled Af ter successful compilation you can again say Runmusc and see the results Please note that if you indeed modified apl_arome F90 you will see no difference with the testset MUTEST because it does not use AROME 7 Handling of MUSC output 7 1 Output variables 7 2 Conversion of LFA files 7 3 Visualisation possibilities 8 Development tasks REFERENCES Boer G J N A McFarlane R Laprise J D Henderson and J P Blanchet 1984 The Canadian Climate Centre spectral atmospheric general circulation model Atmos Ocean 22 397 429 Appendix e gl for extraction from HARMONIE files acadfa pgd prep for academic cases ddh toolbox xmgrace gnuplot e for simp
Download Pdf Manuals
Related Search