Home

HYPROP-FIT Manual

1. 4 PRL HYPROP csv files The subsequent group of entries in the File menu shows recently opened projects allowing a direct and quick access to the projects by double clicking on the project name Close closes the data file If changes have been made a confirmation to save changes is asked for 12 HYPROP FIT Software 3 2 3 Menu Extras Converting and joining files Help Ba Convert tensioVIEW Project tvp files Join files to on iteration file This menu is used to open tensioVIEW projects from HYPROP measurement campaigns and to combine individual measurements to combined projects IN Please select the tensioVIEW file s to export GA measureme Labormethoden 2010 v 5 Labo Organisieren v Neuer Ordner A A A Favoriten Name Anderungsdatum E Desktop _ 2010 Labormethoden_Set3 tvp 03 12 2010 18 44 mi Downloads _ 2010_Labormethoden_Setl tvp 28 11 2010 21 15 E Zuletzt besucht _ _ 2010_Labormethoden_Set2 tvp 26 11 2010 08 30 BB Desktop Bibliotheken Bilder Z5 Dokumente a Musik F Videos B wdurner j Computer amp Windows7_0S 3 Dateiname v tensioVIEW Project tvp z Abbrechen r Ga Ji measureme Labormethoden 2010 v 4 Labormethoden 2010 durchsuc Organisieren v Neuer Ordner E Name nderungsdatum E Desktop EARN N _ Gruppe 5 bhd 03 12 2010 19 41 Bibliotheken SRA
2. 3 Epot Evaluation Method Simplified Evaporation Method GG 10 22 01 2015 11 54 28 10 02 2015 12 04 28 19 Days Geometric Parameters j Zero Offset Correction Typ of sample ring 250 mi i Upper Tensiometer hPa 9 Soil surface area cm2 50 Lower Tensiometer hPa 0 Soil column height cm 5 00 Position lower tensiometer cm 1 25 Position upper tensiometer cm 3 75 Soil volume cm3 249 HYPROP Parameters j Measurement Uncertainty Empty soil sampling ring weight o 201 7 Serial Tensiometer hPa 0 Measurement head net weight a 363 9 Scale o 0 Air entry pressure upper tensiometer bar 8 8 Air entry pressure lower tensiometer bar 8 8 Density of solid substance g cm3 2 65 Notes Site coordinates Sampling date Site description Soil type Land use Soil physical data Texture clay sit sand The Information Window contains nine groups of information blocks about the project Some of the data are editable by the Public User some more are editable by the Power User and others that cannot be altered by the user and are listed as information that is specified by the TensioView or HYPROP VIEW data acquisition program and the respective firmware In a typical measurement evaluation there is one single input value that must be specified by the user in this window This is in group five HYPROP Parameters the Empty
3. yGm uni yes no 3110 a n k e ho m vGm uni no yes 3101 a n wl k r hojwtalm jvem uni yes yes 3111 a n w k t holw alm pf Tt ae pvem bi yes no 1210 a n o K r h o n we m m pvem bi no yes 1201 a n w K t h o a a n we m m yGm bi yes yes 1211 a n w Ks t h o a jo m2 we m m Tl ht o Bc Bur 4000 hs AO Os Kio p ft ft Px juni 5000 a n h ke e hofm FX uni yes yes 5114 a n h K t hofm alm Tl FX bi ves ves 5224 a n he Ks ho a a2 m we m m ft tt tt tt tt tt o 61 HYPROP FIT Software 3 Parameter estimation 3 1 Definition of objective function The parametric models for O h and K h are fitted simultaneously to the data points This is essential as distinct parameters e g and for the van Genuchten Mualem model influence the shape of both functions The fitting is accomplished by a non linear regression algorithm by minimizing the sum of weighted squared residuals between model prediction and data pairs b We Xi 1 Wo i 9 6 b We Die Wei Ki R b 52 where rand k are the number of data pairs for the retention and the conductivity function respectively Wo and wx are the class weights
4. zert ertc tro ge caw i 2hm Kei St 1 Wid J a72 26 Liza Wizna to with P 1Wi5 erf erfc 21rb 4 27 for the bimodal Kosugi function Note that for Io 0 equations 24 27 reduce to equations 13 16 In case of using the unconstrained uni or bimodal van Genuchten function the capillary bundle model of Mualem equation 17 is solved numerically 1 2 3 Film conductivity The relative conductivity for water flow in films is given by Peters 2013 gfim _ e rel ha 28 where a is the slope on the log log scale which is might be fixed to 1 5 Tokunaga 2009 Peters 2013 55 HYPROP FIT Software 1 2 4 Prediction of isothermal vapor conductivity SHYPFIT2 0 allows optional to ad isothermal vapor conductivity to liquid hydraulic conductivity In that case total hydraulic conductivity is given by K K 1 w KOS SP wKIE S4 KP 29 The isothermal vapor conductivity K is calculated according to Saito et al 2006 as M Ge my pai 2 30 Pp RT where px kg m and ow kg m ow 1000 kg m are the saturated vapor density and the liquid density of water M kg molt M 0 018015 kg molt is the molecular weight of water g m s g 9 81 m s is the gravitational acceleration R J molt kgt R 8 314 J molt kg is the universal gas constant T K is the absolute temperature D m s is the vapor diffusivity and H is the relative humidity D is de
5. unimodal bi bimodal Scaling Scal and Adsorption Ads are either considered or not If Ads is considered film conductivity is considered as well Note that the models that are greyed out are not directly implemented in HYPROP FIT ee ee ee ee ee Parameter Basic_ Mod Scal Ads Code 1 2 3 4 5 6 7 8 9 10 11 12 13 14 vG ui no no 1100 a n ofk yG ui yes no 1110 a n ofk frf vG ui no yes 1101 a n wfo K r hjo a vG uni yes yes 1111 a n w fK r hbjl o a o T Kos uni no no 2100 hm o jo frj Kos uni yes no 2110 hm o ofk jrfn Kos uni no yes 2101 hm o w K r hjoja Kos uni_ yes yes 2111 hm o w k t hofm a fT ht vG bi no no 1200 a n 0 K e jo m we yG bi yes no 1210 a n o K e hol o m we yG bi no yes 1201 a n w o K r h o a o m tw vG bi yes yes 1211 a n w K t h o ajo m iw o T T o Kos bi jno no 2200 hm o 0 Ke t f hm2 or we Kos bi yes no 2210 hm o Ks hol mow Kos bi no yes 2201 hm o w ks holw a hm2 o2 we Kos bi__ yes ves 2214 hm o w Ks holm a hm2 o2 we p Tl o pvem uni no no 3100 a n e l h im
6. Closed form expression for the multi modal unsaturated conductivity function Vadose Zone Journal 5 121 124 Romano N P Nasta G Severino and J W Hopmans 2011 Using Bimodal Lognormal Functions to Describe Soil Hydraulic Properties Soil Sci Soc Am J 75 2 468 480 Saito H J Simunek and B P Mohanty 2006 Numerical analysis of coupled water vapor and heat transport in the vadose zone Vadose Zone J 5 784 800 doi 10 2136 vzj2006 0007 Schelle H S C Iden A Peters and W Durner 2010 Analysis of the agreement of soil hydraulic properties obtained from multistep outflow and evaporation methods Vadose Zone Journal 9 1080 1091 doi 10 2136 vzj2010 0050 Simunek J and J Hopmans 2002 Parameter optimization and nonlinear fitting In Dane J H Topp G C Eds Methods of Soil Analysis Part 4 Physical Methods Soil Science Society of America Book Series No 5 Madison Wi pp 139 157 van Genuchten M T 1980 A closed form equation for predicting the hydraulic conductivity of unsaturated soils Soil Sci Soc Am J 44 892 898 Ye M P D Meyer and S P Neuman 2008 On model selection criteria in multimodel analysis Water Resources Research 44 3 W03428 doi 10 1029 2008WRO006803 68
7. G von Unold L Mueller and R Wieland 2010b The evaporation method Extending the measurement range of soil hydraulic properties using the air entry pressure of the ceramic cup J Plant Nutr Soil Sci 173 563 572 Wind G P 1968 Capillary conductivity data estimated by a simple method In Proc UNESCO IASH Symp Water in the unsaturated zone Wageningen The Netherlands 181 191 43 DIB HYPROP FIT Software Appendix 1 Theoretical basics of the Simplified Evaporation Method HYPROP FIT Software In a soil sampling ring two tensiometers comparable to the T5 model are in stalled at two depths z and z2 The middle height between the sensing tips of the tensiometers is the central plane of the soil sample The sample is saturated closed on the bottom and placed on a scale The upper side of the sample is open to atmosphere so that the soil moisture can evaporate From the soil water tension hPa the average matric potential and the hydraulic gradient are calculated The mass difference measured by the scale is used to calculate the volumetric water content and the waters flow rate A measuring campaign will last until one of the tensiometers runs dry or the mass changes become marginal Then the dry weight of the soil is determined by oven drying the sample at 105C for 24 hours With these values the retention curve and the unsaturated conductivity is extrapolated as described next The total weight of water w
8. If individual measurements for the soil under investigations are available this value can be replaced by a measured one The value is used to calculate the porosity of the sample from the dry soil weight It is also used for the automatic estimation of the saturated water content as described in Appendix 2 19 201g Yes 358g Yes 8 8 No bar 8 8 No bar 2 65 No g cm Yes Yes Yes Yes Yes HYPROP FIT Software Measurement Uncertainty Measurer Variable Tensiometer Scale ent Uncertainty Tensiometer hFa 0 15 Scale g 0 05 Explanation default editable editable value by by public power user user Statistical measurement uncertainty of the tension 0 15 No Yes reading i e standard deviation of repeated hPa tensiometer readings at a constant physical pressure head The tensiometer reading uncertainty is used to calculate the parameter uncertainty of the results as indicated by the confidence limits given in register fitting IMPORTANT The accuracy of the tensiometers is related to random fluctuations i e noise Offsets can lead to systematic errors Users must control whether the difference of the tensions under hydrostatic conditions are 2 5 hPa Otherwise erroneous conductivity calculations will occur In that case it is advisable to either correct the tensiometer offset or if this is difficult or impossible to increase the uncertainty of the tensiometer A
9. The model structure in SHYPFIT2 0 follows the PDI model suggestion which combines water retention and conductivity in completely and incompletely filled capillaries as well as isothermal vapor conductivity Peters 2013 Iden and Durner 2014 Peters 2014 The commonly used retention models of van Genuchten 1980 Kosugi 1996 or the bimodal model of Durner 1994 in combination with the Mualem 1976 hydraulic conductivity model are given as special cases of the general PDI models as will be outlined below 49 HYPROP FIT Software 1 PDI model combinations 1 1 Retention functions 1 1 1 General The general form of the PDI retention model is given by the sum of capillary and adsorptive water retention Peters 2013 Iden and Durner 2014 O h 6 6 S 6 54 1 where 9 is the volumetric water content h cm is the suction S and S are the saturation of capillary and adsorptive water retention 9 is saturated water content and 0 is the maximum water content for water adsorption Capillary retention 0 h is given by the first term on the right side of equation 1 and adsorptive retention 62 h by second term Setting S to 1 gives the original capillary retention functions with residual water content A typical unimodal PDI soil water retention function is shown in Fig 1 10 10 10 10 10 10 10 h cm Figure 1 Typical unimodal PDI soil water retention function The
10. _ Gruppe 6 bhd 03 12 2010 19 42 3HD Date Bilder Gruppe 7 bhd 03 12 2010 19 44 Dat L Dokumente y _ Gruppe 8 bhd 18 04 2011 09 06 Datd a Musik _ Gruppe01 bhd 26 11 2010 10 39 Dat Videos Gruppe02 bhd 26 11 2010 10 55 3HD Date B wdurner Gruppe03 bhd 01 12 2010 15 25 Datd jE Computer Gruppe04 bhd 26 11 2010 11 59 Dat amp Windows7_0S Gruppe09 bhd 26 11 2010 12 10 HD Data ees Wechseldatent mi Gruppel0 bhd 26 11 2010 12 19 Datd ES Lenovo_Recove Gruppell bhd 19 04 2011 13 45 3HD Date ublic boku E ANE aE Gruppel2 bhd 26 11 2010 14 20 3HD Date G2 boku_www r G confidential T lt r Dateiname v _ HYPROP binary project files t v a Ld b C 13 Convert tensioView Project tvp files is used to read existing tensioView data files that have been produced by a HYPROP measuring campaign and stored by the data acquisition system in a tensioView project file It opens a browser window from where the files that are to be imported can be selected The imported file will be immediately converted to one or more HYPROP binary data bhdx files and can be stored and further processed Join files to an iteration file is used to combine measuring campaigns that are stored as bhd or bhdx files to one combined project file of type bhdix It opens a browser window from where the files that are to be combined can be selected T
11. a guess of the initial water content can be obtained using the HYPROP FIT software by 0 Mo Pu Vior l 2 Pp Vor where cm3 cm is the saturated volumetric water content equal to the porosity and assumed to be the initial water content pw g cm is the density of water taken to be 1 0 g cm s g cm is the density of the soil matrix Vio cm is the volume of the soil sample and mo g is the mass of water plus soil at the beginning of the experiment 1 e at full saturation The value is taken from the total recorded weight at the start line whereas all other values are as specified in the Information register Please note that this calculation assumes that all pores are water filled so that the porosity is equal to the saturated water content Since this is rarely the case in real evaporation measurements this approximation is only a reasonable initial guess An accurate calculation requires weighing the dry soil and inserting the respective value in the menu Calculation of water contents in the register Evaluation 48 DI HYPROP FIT Software Appendix 3 SHYPFIT2 0 User s Manual Authors Andre Peters and Wolfgang Durner 2015 Cite as Peters A and Durner W 2015 SHYPFIT 2 0 User s Manual Research Report Institut fur Okologie Technische Universitat Berlin Germany HYPROP FIT uses SHYPFIT2 0 for selecting and fitting different soil hydraulic models to the data
12. 53 35 16 01 2015 08 34 44 962 41 412 01 2 28 17 01 2015 09 10 49 953 93 403 53 70 76 Sat 10 Sun 11 Mon 12 Tue 13 Wed 14 Thu 15 Fri 16 Sat 17 Sun 18 Mon 19 Tue 20 Wed 21 Weight change A o AAir Entry Point top Air Entry Point botton Weight g Stop top Stop bott The lower graph is the weight graph It shows a grey line that indicates the net weight of the sample i e the mass of dry soil plus the mass of water The accuracy of the absolute level of these data depends on the accuracy of the values for the measurement head and the steel cylinder and possibly additional weights as described in the register Information The accuracy of the weight changes i e the relative accuracy is equal to the accuracy of the used scale The times of the Stop and the Air entries are shown by vertical lines at identical positions to the tension graph The x and y axes are automatically scaled to show all recorded data Parts of the graph can be manually enlarged and reset as described for the tension graph Note that changing the time axis of the weight of one graph does not affect the tension graph and vice versa Visualization of the actually measured points and the spline support points can be activated and de activated as described for the tension graph Interpretation of the weight data In a typical measurement the weight data will first show a continuous weight loss with an almost constant rate whic
13. Effectively the measured mean water content of the column is the integral of the water content distribution over the whole column divided by the volume of the column With the assumption of hydraulic equilibrium for hydrostatic experiments or quasi equilibrium for the early stage of evaporation experiments Peters and Durner 2008b this is in the one dimensional case the same as the integral of the retention function over the matric heads from the lower boundary of the column to the upper boundary divided by the height of the column Therefore we have to replace in Eq 52 the model predicted point water content 6 b by the model predicted mean water content xz hy iz 6 b 7 J b h dh 53 where L cm is the column height and hub and hip indicate the upper boundary and lower boundary suctions for each equilibrium state If the option integral fit is selected the column length must be given In this case SHYPFIT2 0 interprets the measured suction as the suction in the soil center and calculates the hu and hu by assuming hydrostatic equilibrium i e h h L 2 and h h L 2 3 3 Weighing schemes Since the objective function b Eq 52 involves data of different types with different measurement frequency the result of the optimization will likely be affected by the weights of the data Simunek and Hopmans 2002 Therefore the user can chose between 3 different weighting schemes In the first scheme the use
14. For small values for n i e wide pore size distributions the saturation of unscaled function does not reach a value of 0 at oven dryness at h ho unscaled scaled 0 8 c Oo 0 6 5 w u v 0 4 D D 0 2 0 0 1 2 3 4 40 a 10 10 10 10 10 10 10 h cm Figure 2 Effect of scaling the capillary retention function according to equation 3 using the van Genuchten function as basic function Unscaled is the original S h I h function and scaled means S h I h T0 1 I0 Numbers indicate different values for n The basic functions implemented in SHYPFIT2 0 are the constrained and unconstrained unimodal function of van Genuchten 1980 the unimodal function of Kosugi 1996 as well as the bimodal form of them Durner 1994 Romano et al 2011 The constrained function of van Genuchten 1980 is given by rh j7 5 1 ah 51 HYPROP FIT Software where a 1 cm and n are curve shape parameters The unconstrained function of van Genuchten 1980 is given by rn 6 1 ah where m is an additionally shape parameter The unimodal Kosugi retention function is given by te 7 Ch ert io where hnm is the suction corresponding to the median pore radius o is the standard deviation of the log transformed pore size distribution density function and erfc denotes the complementary error function The bimodal functions are weighted sums of the unimodal
15. deletion of data will be immediately shown in the graph to the left 4 2 7 The weight data window The window at the lower right shows the recorded weight data After date and time column 1 the total weight of the measuring system including all tare weights is listed column 2 The net weight of soil plus water is calculated from the total weight minus the tare weights as specified in the Information register and listed in column 3 These data are also depicted on the graph to the left Finally the net weight change from the start of the experiment is listed in column 4 The data window can be scrolled by the scroll bar to the right By moving the cursor into the data window and clicking on the right hand mouse button the data can be directly edited by Power Users Possible operations are deletion of lines insertion of lines and changes of the contents of individual data cells Also multiple lines can be selected and the data can be copied with the ctr c option Manipulation of data or deletion of data will be immediately shown in the graph to the left 28 HYPROP FIT Software 4 3 Register Evaluation Calculating retention and conductivity data al HYPROP FIT C Users wdurner Dropbox Daten _projektbezogene Daten Laborpraktikum 2014 HYPROP HYPROP 3 Schunter 07 bhdx 3 1 12 16476 H File Extra Help www ums muc de im ral Information EA Measurements Ad Evaluation Fa Fitting 3 Export Calculation of water co
16. discussion of principles Journal of Hydrology 10 282 290 Omlin M and P Reichert 1999 A comparison of techniques for the estimation of model prediction uncertainty Ecol Model 115 45 59 Peters A 2013 Simple consistent models for water retention and hydraulic conductivity in the complete moisture range Water Resour Res 49 6765 6780 Peters A 2014 Reply to comment by S Iden and W Durner on Simple consistent models for water retention and hydraulic conductivity in the complete moisture range Water Resour Res 50 7535 7539 Peters A and W Durner 2006 Improved estimation of soil water retention characteristics from hydrostatic column experiments Water Resour Res 42 W11401 doi 10 1029 2006WRO004952 Peters A and W Durner 2008a Simplified Evaporation Method for Determining Soil Hydraulic Properties Journal of Hydrology 356 147 162 doi 10 1016 j jhydrol 2008 04 016 67 Ds HYPROP FIT Software Peters A and W Durner 2008b A simple model for describing hydraulic conductivity in unsaturated porous media accounting for film and capillary flow Water Resour Res 44 W11417 doi 101029 2008WRO007136 Peters A S C Iden und W Durner 2015 Revisiting the simplified evaporation method Identification of hydraulic functions considering vapor film and corner flow Journal of Hydrology 527 531 542 doi http dx doi org 10 1016 j jhydrol 2015 05 020 Priesack E and W Durner 2006
17. for the functional relationships 66 HYPROP FIT Software 5 References Akaike H 1974 A new look at statistical model identification IEEE Trans Autom Control AC 19 716 723 Brooks R H and A T Corey 1964 Hydraulic properties of porous media Hydrol Paper 3 1 27 Colorado State Univ Fort Collins Duan Q S Sorooshian and V Gupta 1992 Effective and efficient global optimization for conceptual rainfall runoff models Water Resour Res 28 1015 1031 Durner W 1994 Hydraulic conductivity estimation for soils with heteroge neous pore structure Water Resour Res 30 211 223 Hurvich C and C Tsai 1989 Regression and time series model selection in small samples Biometrika 76 2 297 307 doi 10 1093 biomet 76 2 297 Iden S and W Durner 2014 Comment to Simple consistent models for water retention and hydraulic conductivity in the complete moisture range by A Peters Water Resour Res 50 7530 7534 Kosugi K 1996 Lognormal distribution model for unsaturated soil hydraulic properties Water Resour Res 32 2697 2703 Marquardt D 1963 An algorithm for least squares estimation of nonlinear parameters J Soc Ind Appl Math 11 431 441 Mualem Y 1976 A new model for predicting the hydraulic conductivity of unsaturated porous media Water Resour Res 12 3 513 521 Nash J E and J V Sutcliffe 1970 River flow forecasting through conceptual models 1 a
18. full days Weight and Volume Correction Comection Volume comection ml 0 Weight comection g 0 00 Volume Correction of the sample volume This can be necessary Ocm No Yes Correction e g if the prepared sample surface is not perfectly prepared and shows deficits negative correction Volume correction will affect the calculation of the water contents since the weight changes will be related to the value Volume Volume correction Weight Addition positive or subtraction negative of aconstant Og Yes Yes correction weight to the Tare weights which are subtracted from the total weight to obtain the net sample weight This could be the constant weight of e g a lid a cable or another sensor that might be added in a measurement campaign 17 HYPROP FIT Software Geometric Parameters Geometric Parameters Typ of sample ring 220 mi ha Soil suface area em2 50 Soil column height em 5 00 Position lower tensiometer cm 1 25 Position upper tensiometer em 3 75 Soil volume cem3 249 Variable Explanation default ed by ed by value public power user user Type of There are two standard types of sample ring 250 cm 250cm Yes Yes sample ring and 100 cm The selectin of this entry will affect the subsequent default entries for surface area and height Soil surface Surface are soil sample Default value depends on 50 cm No Yes area selection of cylinder type Soil column Height of s
19. has been determined after the end of the measurement campaign by oven drying at 105 C for 24 hours The exactness of the results depends on the correctness of all involved parameters that are specified in the Informations register From the data further soil physical parameters such as the initial water content the identical porosity the dry bulk density and the dry soil weight are also calculated and shown in the underlying group Calculated Parameters IMPORTANT Only option 1 leads to a correct calculation of the absolute water contents Option 2 External setting of the inital water content by user Calculation of water contents From dry soil weight g From given initial water content olz Estimation of initial water content This option allows a user to externally set the initial water content to a desired value from wherever he takes this value Since the saturation state associated with the externally set initial water content is unknown no calculation of porosity dry bulk density and dry soil weight is possible and no values of these parameters can be shown Option 3 Automatic estimation of the inital water content Calculation of water contents From dry soil weight g From given initial water content ol Estimation of initial water content Calculated Parameters Initial water content Vole 38 2 Dry sol weight g 407 9 Density g cm 1 64 Porosity 0 38 If the dry wei
20. of the water content data and conductivity data wo and wx are the weights of the individual data points and 6 0 b Ki and K b are the measured and model predicted values respectively and b is the parameter vector In SHYPFIT2 0 either the combination of O h and K h models or only one of them can be fitted to the measured data In the latter case the first or second part of Eq 52 vanishes The parameter estimation procedure must guarantee that the best parameter combination for the appropriate model combination is found i e the global minimum in the multidimensional parameter space must be determined Often local estimation algorithms such as the Levenberg Marquardt algorithm Marquardt 1963 are used for parameter estimation However especially for complex models with many degrees of freedom these algorithms fail to find the optimal parameter vector so that the outcome of these algorithms largely depends on the initial guesses Therefore SHYPFIT2 0 uses the shuffled complex evolution algorithm SCE Duan et al 1992 which is a global parameter estimation algorithms One crucial part is the choice of the correct boundaries for the parameters In SHYPFIT2 0 the boundaries of the parameters for each model combination have default values which might be changed by the user if necessary The predefined boundaries are chosen in a way which shall guarantee high flexibility on the one hand and physical consistency on the other hand To gua
21. resembles the classic soil water retention function with residual water content 1 2 Conductivity functions 1 2 1 General To be valid for the complete moisture range the liquid hydraulic conductivity model is given by a sum of capillary and film conductivity Peters 2013 K K 1 w KS S wK S 24 11 rel rel Koo vr H and K4 are relative conductivities where K cm d is the saturated liquid conductivity for capillary and film flow 1 2 2 Capillary conductivity Relative conductivity for capillary flow is described by the pore bundle model of Mualem The Mualem model for the capillary retention function is given by h 11 7 6Ca fo zas p ca 2 Prase En 12 erem eer where 7 is the tortuosity and connectivity parameter and S is a dummy variable of integration Except for the unconstrained van Genuchten variants there exist analytical solution for all above mentioned soil water retention functions 1 2 3 1 Analytical solutions for unscaled capillary retention functions For the unscaled capillary retention functions S I In these cases the analytical solutions are given by 2 KOP Soo 1 _ 1 Z seapyt my 13 rel 53 HYPROP FIT Software for the constrained van Genuchten function 2 Ko So erfc erfc 25 z 14 for the Kosugi function E Wii i 1 ace alas i 1 K o Wi sear 1 i 15 For the bimodal van Genuchten function and o
22. soil sampling ring weight g Since the sampling ring weight differs from sample to sample the weight of the ring must be specified for each individual measurement In the following the data groups will be listed and the inputs will be explained Evaluation Method Evaluation Method Currently the Simplified Evaporation Method is Simplified Evaporation Method implemented Schindler 1980 Peters et al 2015 16 HYPROP FIT Software General Information General Information Sample name GG 10 Start of measurement 22 01 2015 11 54 28 Stop of measurement 10 02 2015 12 04 28 Duration of Measurements 19 Days Variable Explanation default Ed by Ed by value public power user user Sample Denomination of a sample The name can be chosen From Yes Yes Name freely by the user It should be noted that the sample HYPROP name is used for the creation of the filename during campaig export by default The filename rules have to be n considered If data stem from a HYPROP measurement campaign the sample name specified in HYPROP VIEW is listed here Start of Date and time from the start of the recorded data No No measure These data are given by the tvp file ment Stop of Date and time from the end of the recorded data No No measure These data are given by the tvp file ment Duration of Difference between end and start date of the recorded No No measure data calculated from the two fields above and ments rounded to
23. tensiometers vertical lines will be shown If the drop of the tension is not clearly indicated by the 26 HYPROP FIT Software course of the data the automatic detection of the air entry point will fail The vertical air entry line is then at a wrong position but can be moved with the cursor as described above for the stop point position After manual shifting the original position due to the automatic detection can be again recovered by pressing Search Air Entry Point When the Use Air Entry Point option is active dashed lines are shown that interpolate the tensiometric data between the last reliably measured points stop point and the air entry points of both tensiometers For the HYPROP FIT calculations the tensiometric value of the lower tensiometer will be taken from this interpolation For further information about using this option the user is referred to the publication of Schindler et al 2010b 4 2 5 The weight graph Weight weight Net weight o 1 09 01 2015 11 07 39 1024694742900 09 01 2015 12 02 20 1024 50 474 10 0 19 09 01 2015 13 3233 1023 73 473 33 096 10 01 2015 0949 49 1016 30 465 90 39 11 01 2015 13 18 39 1006 35 455 95 18 34 12 01 2015 08 45 36 998 44 448 04 26 25 12 01 2015 19 30 02 994 19 443 79 30 50 13 01 2015 08 05 10 989 76 439 36 34 93 14 01 2015 08 17 02 980 91 43051 43 78 14 01 2015 18 12 49 976 86 426 46 4783 15 01 2015 08 00 14 971 34 420 94
24. that can be directly evaluated This stop point is automatically detected by the software and indicated by a vertical dashed line The stop line can be manually shifted to an earlier or later time be placing the mouse cursor on the stop line and dragging it to the left or to the right The automatic detection of the stop point can again be applied by clicking on the Search Stop Point button immediately above the graph If tensiometers were not well prepared the boiling retardation is not reached and the sudden collapse is not visible but replaced by an asymptotic approach of the tension towards a final value In these cases the automatic stop point detection is not fully reliable and users must carefully check the end position As a rule of thumb the stop point must occur at a point where the slope of the tension curve is still increasing This is illustrated in the figure below File Extra Help Olh O GJ trfomation FO Measurements Fa Evauation o Fiting X Epor Search Stop Point Use Air Entry Point Search Air Entry Point Show Spline Interpolation Points Show Data Points Tension bottom Tension top Tension bottom Tension top Temperature hPa hPa C 24 11 2014 08 46 54 24 11 2014 08 48 54 24 11 2014 08 49 54 24 11 2014 08 50 54 24 11 2014 08 51 54 24 11 2014 08 52 54 24 11 2014 08 53 54 24 11 2014 08 54 54 24 11 2014 08 55 54 24 11 2014 08 56 54 24 11 2014 08 57 54 24 11 2014 08 58 54 24 11 2014 0
25. to combined data sets in so called HYPROP binary iteration projects files with extension bhdix This is of particular use if data sets are to be compared or if multiple data sets are to be fitted with a single hydraulic function The process to combine the individual bhdx files to a bhdix project is described the Extras Command menu section 3 2 3 Once individual measurement campaigns are combined they can be activated and de activated simply by clicking on the individual tick boxes in the file explorer As an example in the figure above five the measurement campaigns of a JKI series are combined but only the first four are activated Data from different campaigns are shown in different colors Fitting of functions occurs simultaneously for all depicted data Individual data sets can be fully edited and processed by single cl icking on the respective campaign name in the file explorer The source file for the data will be unaffected i e the bhdi file contains copies of the original data Clicking on the bhdix name will return to the combined project menu as shown above 42 HYPROP FIT Software 6 References Akaike H 1974 A new look at statistical model identification IEEE Trans Autom Control AC 19 716 723 Peters A and W Durner 2006 Improved estimation of soil water retention characteristics from hydrostatic column experiments Water Resour Res 42 W11401 Pete
26. to the results Furthermore overfilling or underfilling of the steel cylinders with soil will cause appreciable errors As a rule of thumb 1 mm error in sample height causes about 2 error in the calculated data 4 3 3 Adding independent retention and conductivity data Calculation of water contents Add retention data points Add conductivity data points Add WP4 data points From dry soil weight g 387 5 m E _ From given initial water content Vol oF kanea Weighting pFH log10 K Weighting pFH z Pe Gross dry ens Weighting ae jisra nt _ Estimation of initial water content Vol factor em d factor Ig mass o Ig factor Mass Vol Calculated Parameters Initial water content Vol 345 Dry soil weight o 387 5 Density g cm3 1 56 Porosity 0 41 Users can add independently measured data e g from WP4 measurements or from pressure plate measurements This is done by typing the value in the respective field To complete the insertion the cursor must be positioned out of the field which can be done by pressing the Tab or Return key or by moving the mouse to another field The additional data are shown immediately in the data graphs by solid markers whereas the HYPROP calculated data are shown as empty circles IMPORTANT Please care for the appropriate units when adding the data tensions must be added as pF water contents in percent and conductivity data as decimal log of the value expresse
27. 2 J tau 0 355 1 677 0 967 iici Perl Frediund Xing O O thor 0 000 0 399 0 399 cm cm v omega 1 09E 9 0 00E 0 _ Infinity WC 6kPa 324 E WC 33kPa 18 3 imizati Kosugi C C C C viith_s 0 333 0 332 0 335 cm cm a 1 500 We 1500 kPa 5 emer eee C T T T eee 0 80 PAW 6_1500kPa 27 4 alpha2 0 0103 0 0099 0 0109 1 cm PAW 33 1500kPa 13 3 van Genuchten mnvar C C C c J n2 7 497 6680 8414 Model FXbi PDI Model Code 6211 7 m2 0638 kiA Sasoi mk ya walla J m1 0 939 6 082 7 960 Description PDI variant of a bimodal Fredlund Xing model v m2 0 177 0 181 0 535 Show data as small points Retention O pF Conductivity K pF Conductivity K O function data function data added data function data added data 35 30 N on vol water content log 10 K in cmid log 10 K in cmd 0 5 10 15 20 25 30 35 vol water content PowerUser X For use in numerical modelling hydraulic characteristics are described by parametric functions for O h and K h or K 0 where is the volumetric water content K the hydraulic conductivity and h the matric potential HYPROP FIT provides a high class algorithm to fit a wide variety of functional relationships of the retention curve and the conductivity curve to the data No specification of initial guesses for the parameter values is requried Five basic types of retention models are available encompassing the expressions of Br
28. 5 13 18 39 Tension bottom Tension top hPa hPa 1 92 1 96 1 98 1 99 2 03 2 10 2 16 2 10 1 66 1 69 1 70 1 72 1 74 1 76 1 78 1 79 315 3 18 3 18 3 19 3 26 3 32 Temperature fC 18 00 18 00 18 01 18 02 18 03 18 04 18 04 18 05 18 05 18 05 18 06 18 07 18 08 18 09 18 10 18 11 Gross weight Weight change A iol Net weight ol i 1024 69 1024 50 1023 73 1016 30 1006 35 474 29 474 10 473 33 465 90 455 95 0 00 0 19 0 96 8 39 18 34 12 01 2015 08 45 36 998 44 448 04 26 25 Sat 10 Sun 11 Mon 12 Tue 13 Wed 14 Thu 15 Fri 16 Sat 17 Sun 18 Wed 21 12 01 2015 19 30 02 994 19 443 79 30 50 Jan 2015 Time Mon 19 Tue 20 PowerUser X The register Measurements shows the readings of the two tensiometers and the recorded weights The register is composed of four windows Two of the windows show graphs of the recorded data two others the numerical values The size of the windows can be freely scaled by positioning the mouse on the invisible borders of an individual window and dragging the border to a new position If the mouse is directly on a border its appearance changes from the arrow to a double arrow Each of the graphics windows can be enlarged to full screen by positioning the mouse on the window clicking the right hand mouse button and selecting the respective option in the pop up menu Leaving the full screen mode is reached by pressing the Escape key A user can repe
29. 60 Where Ji w is the determinant of the Fisher information matrix J is the Jacobian or sensitivity matrix with elements J 0Z 0b where Z is the model prediction at point i and bj is the j th parameter The superscript T indicates the transpose of the matrix and w is the weighting matrix 4 2 Uncertainty analysis 4 2 1 Parameter uncertainties In order to evaluate the uncertainties of the estimated parameters a linear approximation of the covariance matrix of the estimated parameters C is calculated Omlin and Reichert 1999 va Din T 1 C bime J wJ 61 The linear approximation leads to a confidence interval for the i th estimated parameterb by bi Cpi tn k a 2 62 where where tn k a 2 is the upper a 2 quantile of the students t distribution with n k degrees of freedom a is set to 0 05 leading to the 95 confidence interval for the parameters 4 2 2 Function uncertainties The uncertainties for the O h and K h functions are determined by calculating the covariance of the model response C C J C J 63 where the J is the Jacobian or sensitivity matrix with elements J 02 0b where Z is the model prediction at point i and bj is the jth parameter The confidence interval for the model output is given by Vj Cz tn k a 2 64 where f is the model output and log K for the retention and conductivity function respectively Again a is set to 0 05 leading to the 95 confidence intervals
30. 8 59 54 24 11 2014 09 00 54 24 11 2014 09 01 54 Tue 25 24 11 2014 09 02 54 Nov 2014 a ea ae T a a 5 Weight 24 11 2014 08 48 05 1021 61 24 11 2014 10 51 14 1020 45 25 11 2014 08 47 37 1009 81 25 11 2014 18 23 28 1005 25 26 11 2014 09 39 36 998 14 27 11 2014 07 58 18 987 15 Tue 25 Wed 26 Thu 27 Fri 28 28112014 16 5202 97148 Weight g S 8 8 Nov 2014 PowerUser a After a certain time with further evaporation the tensiometer cups will dry out and become permeable to air At that time the absolute pressure inside the tensiometers will increase from the vapor pressure to the pressure of the ambient atmosphere Accordingly the recorded tension pressure difference between internal water pressure and ambient atmospheric pressure will go down to zero 4 2 2 Data Visualization Options N Search Stop Point Use Air Entry Point Search Air Entry Point Show Spline Support Points Show Data Points Reset Update In the header line above the graphs buttons for data visualization options are listed They differ for power users and public users The options are as follows 24 HYPROP FIT Software Variable Explanation default ed by ed by value public power user user Search Stop The stop point of the measurement is reached by Automatic No Yes Point the cavitation of the upper tensiometer which is indicated by a sudden pressure drop HYPROP FIT detects this point autom
31. Data VisualizatioMm OPON S sierras R om cacta anny snctaeeeatee sea recactusene 24 Ao WINE en OnE AO ee E E 25 4 2 4 Using the air entry point of the tensiometer cup as additional tension measurement Power SIG serria E T O 26 Ao TACNE ENCE DI eoa E ne ee ee eee 27 4 2 6 The tension data windoOW ssssssssesesssssssseresssssssrerressssssterressssssrrrreesssssrrrrressssseeeresessseeeereee 28 4 2 7 The weieht data WINGOW Aen nen ee ee eee neo aa 28 4 3 Register Evaluation Calculating retention and conductivity data cccccceecscseesseeeeseeeeeeens 29 4 3 1 Calculation of retention and conductivity CAtA c cc eeeccccessececeesecceceeseceeeeeceeeeeeceeseeeeeeees 29 4 3 2 Calculation of absolute Water contents ccccccccccccssessseeecceceeseeesseeecceeessaaeasseseceeeesanaeseess 29 4 3 3 Adding independent retention and conductivity Cata ceccceeccseecceeceeeeeceeceeeeeeeeeeeeneees 31 4 3 4 Modifying evaluation parameters Power User Only ccccccccsssececeeeesecceeseeeseeeessaeeneeess 32 4 4 Register Fitting Fitting hydraulic FUNCTIONS tO the data ccceesccccsseceeeeseeceeeeseceeseeeceeeeees 33 AA EPDE CC TO tetas eave nsnciee E cedars seucunosetue set cevesseendand acne aoe 33 4 4 2 Selecting a model and specifying parameter DOUNAS ccccccesseccecesseceeeeseceeeeeeceeeeeeeeeeas 34 4 4 3 Results of the parameter estimation and uncertainty measures ccccccesseceeeeee
32. HYPROP FIT Software al HYPROP FIT HYPROP FIT User s Manual Version 3 0 June 2015 HYPROP FIT Software DISCLAIMER This manual documents the use of the HYPROP FIT Software a computer program for analyzing data from evaporation experiments and fitting the unsaturated soil hydraulic properties HYPROP FIT Software is a public domain product and may be used and copied freely The code has been tested against a large number of soil hydraulic data sets and was found to work correctly However no warranty is given that the program is completely error free If you do encounter problems with the code find errors or have suggestions for improvement please contact Wolfgang Durner or Thomas Pertassek Soil Physics UMS GmbH TU Braunschweig Gmunder Str 37 38106 Braunschweig D 81379 Munchen w durner tu bs de tp ums muc de Tel 49 0 531 391 5605 Tel 49 0 89 12 6652 CITE AS Pertassek T A Peters and W Durner 2015 HYPROP FIT Software User s Manual V 3 0 UMS GmbH Gmunder Str 37 81379 Munchen Germany 66 pp THIS MANUAL INCLUDES THE SHYPFIT 2 0 USER MANUAL AS APPENDIX 3 TO BE CITED AS Peters A and Durner W 2015 SHYPFIT 2 0 User s Manual Research Report Institut f r kologie Technische Universitat Berlin Germany HYPROP FIT Software Content 1 Overview and Scientific Background cccccesseccccseccccesececeuececceecceeeusececeueceeeeeeceeseuecessenecetseges 6 2 NAG VaNAEIOI OF C
33. Interpol is set to 0 then the degree of the polynomial used Tensiometer_mean Determines the calculation mode for the mean tension for the calculation of the water retention curve points 1 arithmetic 2 geometric 3 harmonic Tpower Determines the spacing of the support points for the spline interpolation A vlaue of 1 means equal spacing in time a value of 0 5 means a Spacing that is proportional to the square root of time Recommended values are between 0 5 and 1 32 default value 100 0 75 editable editable by public user No No No No No by power user Yes Yes Yes Yes Yes HYPROP FIT Software 4 4 Register Fitting Fitting hydraulic functions to the data w HYPROP FIT C Users wdurner Dropbox Daten _projektbezogene Daten Laborpraktikum 2014 HYPROP HYPROP 3 JKI 07 bhdx 3 1 12 16476 gt El File Extra Help www ums muc de O Infomation ea Measurements Aa Evaluation a Fitting D Export Soil hydraulic model selection Retention function parameters Conductivity function parameters Statistical analysis bimodal Parameter Vale 25 975 Unt Parameter Vale 25 975 Unt OD d original PDI bimodal pp beeen a a i i oa RMSEK 0 0207 Correlation Matrix alphat 0 00283 0 0000 0 2465 1 cm Iv ks 22 7 0 0 762012 cm d AlCe 1519 Parameter Brooks Corey C I el J ni 1 010 0 492 207
34. NS SOG EW AS E E E E E E E E teusaee reves 8 2 1 Installation using the Microsoft Installer cccccccssseccceesecceceesececeeseceeeeeeceeeensecesseecessuneceesenes 8 2 2 Access rights Public User and Power USE cceeccccessececcesecccceesececeueceeeeeeceeeensecessuaeceeeugeceeseees 8 3 Starting the Software and Loading a Proje ct ccsscccssssesecsonssectonssscuesssessoussestonssscueussscneesssenss 9 Sl Ue CNS SN acc cresteettieatte ca cee ioe dene ctengeiecenae sea E 9 3 2 Opening processing and storing projects the commands File Extras Help ccccceeeeee eens 10 3 2 1 Reading HYPROP VIEW project data from old HYPROP measurement campaigns 11 3 2 2 Menu File Opening modifying and storing HYPROP projects ccccccsssececesseeeeeeeeees 11 3 2 3 Menu Extras Converting and joining files cccccccsscccceessececeesececeeececeeeeeeeseesecetseneess 13 3 2 4 Menu Help HYPROP FIT Version INfOrMatiOn c ccseccccssseccceesececeesececeeeeecessuseeetsuneess 14 i POC SSSI ea ea E E nag boi euseciei tatu N E 15 4 1 Register Information Specifying the required parameters for the measurement 16 4 2 Register Measurements Visualization and editing of measured data cccccsesecceesseeeeeeees 23 4 2 1 Interpretation Of the tensioMetric data cccccccsssccccssscccceesececcenecceceeeceeeeseceeeeeceeseeeeeetas 23 4 22
35. Parameters and Default Public Power value User User can can change change value value 500 No Yes 3 No Yes 7 No Yes HYPROP FIT Software Show_Uncertainty _ Bands WEIGHT_FLAG wk wtheta Flag for showing the 95 uncertainty bands False No Yes for the fitted functions in the depicted graphs by a grey shading The simultaneous fit hydraulic functions to False No Yes retention data and conductivity data is a multi objective minimization problem which requires weights for the data of the different data types HYPROP FIT provides the possibility to either pre scribe these weights WEIGHT_FLAG False or else to let the software iteratively balance out the weights in a way that both data sources retention points and conductivity points contribute in the same order of magnitude to the total fitting error WEIGHT_FLAG True Since the retention data are normally better determined and cover a wider moisture range we recommend to use the default values where the weighting of the conductivity data is enhanced by a factor of 10 Weight of the conductivity data class Default 0 001 No Yes 0 001 if WEIGHT_FLAG false Weight of the retention data class Default 1 0 No Yes 1 0 if WEIGHT_FLAG false 38 HYPROP FIT Software 4 5 Register Export Exporting data graphs functions and parameters File Help D EA Infomation TA Measurements Ra Evaluation Fa Fitting F Export z D File Nam
36. User English If previous data sets have been loaded HYPROP FIT will automatically load the last used data set If the last exit of the program was not regular e g a crash caused by a corrupted data file the program will again appear with an empty screen The welcome screen shows three header lines 1 Atop header line embedded in the blue window frame which indicates the program name with the project name in brackets left and the maximize minimize and exit button flushed right 2 A command line which contains three commands File Extras Help left and a link to the UMS website right 3 An icon line with icons to open the File explorer and a Open file and Save file button Additionally the User level is depicted to the right of the bottom line The settings can be altered by selecting the respective option in the draw down menu that opens when clicking on the vertical triangle HYPROP FIT Software 3 2 Opening processing and storing projects the commands File Extras Help File Extra Help Ee Open Ctri O Save Ctri Ey Saveas Import 1 Ce GGx bhdix 2 C GG6 10 bhdx 3 C GG 09 bhdx 40 GG6 08 bhdx 9 C GG6 07 bhdx 6 C4 G6 06 bhdx 7 C GG6 05 bhdx C GG6 04 bhdx 9 C GG6 03 bhdx 10 CA 406 02 bhdx 11 CA G6 01 bhich 12 C Demol bhd Close All command uses follow the standard Windows System conventions as k
37. Yes tensiometer tensiometer Lower Added offset for the measured values of the lower OhPa No Yes tensiometer tensiometer Sensor Unit Information Sensor Unit Information Note the information in this parameter group is Serial number 0058 not affecting the calculations values are given sensor unit name No 38 purely for information purposes Busnumber 1 Subaddress 5 Tensiometer Version NaN Firmware Version NaN Variable Explanation default editable editable value by by public power user user Serial Information given by UMS firmware Please note this No No number number is always a unique identification of the measurement device Together with the date information it is a unique identifier of the measurement campaign Sensor unit Information as specified during data acquisition No No name Busnumber Information as specified in the data acquisition setup No No Subaddress Information given by UMS firmware As above No No Tensiometer Information given by UMS firmware No No Version Firmware Information given by UMS firmware No No Version 21 HYPROP FIT Software Scale Information scale Information Note the information in this parameter group is a not affecting the calculations values are given Scale name purely for information purposes Serial Information given by UMS firmware through tvp No No number file Scale name Information as specified in data acquisition and No No stored in th
38. ameters vi Export of graphs raw data calculated data fitted functions and other parameters of interest The evaluation of HYPROP measurement campaigns follows the simplified evaporation method This method uses weight changes of samples and the matric potential measurements in the samples during a drying process caused by evaporation to derive soil hydraulic functions The principles of the method were first formulated by Wind 1968 A simplification of the Wind method which is implemented in the HYPROP measurements was proposed by Schindler 1980 Further advancements of the method are included in HYPROP FIT such as the extension of the measurement range towards higher suctions by using tensiometers with boiling retardation Schindler et al 2010a and a further extension of the measurement range by considering the air entry value of the porous cups of the tensiometers Schindler et al 2010b The validity of the simplified evaporation method has been investigated by Peters and Durner 2008 and Peters et al 2015 The detailed implementation of the method in the HYPROP FIT software follows the description given in these papers It includes features that lead to an optimization of the method such as a corrected fit of the hydraulic functions by the integral method to avoid bias in hydraulic properties near saturation Peters and Durner 2006 an Hermitian spline interpolation to the raw measured data to obtain smooth and continuo
39. and _COND csv or in a project file format fit or fitx The x extension indicates files that are processed with HYPROP FIT 3 0 and higher This new software version uses a slightly extended file format as compared to the previous HYPROP Fit 2 0 versions There is upward compatibility This means that old projects e g bhd can be opened with HYPROP 3 0 but are then stored in the new bhdx format It is not possible to open the bhdx files with the previous versions of the HYPROP FIT software The same applies to the bhdi and fit files 10 HYPROP FIT Software 3 2 1 Reading HYPROP VIEW project data from old HYPROP measurement campaigns Convert tensioVIEW Project typ files Join files to on iteration file Preferences Ordner suchen Please select the target directory 47 Computer amp Windows7_Q C d ey Wechseldatentrager D gt 11 MUSIS Lauterbad d 2011 04 14 GirlsDay2011 gt di current GEOFLUXES gt HYPROP To import raw data from measurement campaigns that were performed with older versions of HYPROP VIEW at that time called tensio VIEW open the submenu Extras and select the option Convert tensio VIEW Project tvp files The imported raw data will be automatically converted to HYPROP binary projects of type bhdx After the successful import immediately the selection of a target directory for the storage of the bhdx projects will b
40. at time t is derived from the weight measurements at time t by subtracting the weights of the tare components from the total weight Wtot i recorded by the HYPROP balance at times t The tare weights in a standard HY PROP mesurement campaign are the weight of the the empty steel cylinder Weyl the weight of the HY PORP base including the tensiometer shafts and the silicon mat wuHy p and the weight of the dry soil wsoi Thus w wtot t Weyl WHY P Wsoil Dividing w by the density of water which is assumed to be equal to 1 g cm7 gives the water volume V The mean water content of the column 6 at time t is computed by dividing the volume of water by the column volume V Note that 6 is calculated for the N data points obtained from the cubic hermite interpolation of the Hydrus 1D output V 500 by default To obtain data of the retention curve each value of 9 is related to a mean suction at time t In the original data evaluation the arithmetic mean of the two measured suctions is used i e h i hii h3 However investigations of Peters et al 2015 have shown that the geometric mean generally is preferable i e h J hii hi HYPROP offers the possibility to select the mean according to user specifications with the geometric mean as default 45 IS HYPROP FIT Software For the calculation of unsaturated hydraulic conductivity one first estimates the 1 a i t it is assumed to water flux at the c
41. atedly magnify portions of the graphs regardless whether in windows of full screen mode by placing the mouse on the graph and drawing the cursor to the opposite point of a rectangle with the left mouse button held down To reset the window ranges click the right hand button of the mouse and select the respective option 4 2 1 Interpretation of the tensiometric data In a long measurement period as shown in the example screen we see after an initial reaction and a plateau phase an exponential increase of both tensions steeper for the upper tensiometer which is followed by a sudden collapse to smaller values This collapse occurs upon cavitation of the water phase in the tensiometer Before this cavitation the water in the tensiometer is due to boiling retardation under a strain i e under negative absolute pressure When the tension is so high that absolute internal pressure is below vacuum the liquid phase at some point suddenly ruptures and a small bubble of vapor arises This causes the pressure inside the tensiometer to suddenly jump from a negative value to the vapor pressure of water To measure this sharp cavitation point at tensions gt 1000 hPa tensiometers must be well prepared without any gas bubbles inside as indicated in the HYPROP manual The recorded tension values are absolutely reliable up to the cavitation point Accordingly 23 HYPROP FIT Software the cavitation point of the upper tensiometer marks the end of the data
42. atically If the Stop line is moved manually to another time it will be reset to the automatic detection by pressing the Search Stop Point button again If the tensiometer is not filled correctly the drop might not occur in that sharp manner since the tensiometer does not reach a tension In that case it might be necessary that a user chooses the Stop point manually A good choice is then to select the point with the highest slope of the upper tension Use Air Activates and deactivates the use of the air entry No No Yes Entry Point point of the porous cup as additional tension value as described in section 4 2 4 Search Air If the User Air Entry Point is activated and the Entry Point line that indicates the time of the air entry point has been moved manually pressing this button resets the automatic detection of the air entry time Show Spline Activates and de activates the depiction the Support supporting data points as described in section Points 4 2 3 Show Data Activates and de activates the depiction of the No Yes Yes Points individual measured data points as described in section 4 2 3 Reset The option Reset resets all settings to default Yes Yes values in particular the ranges for the visualization of the data in the two graphs 4 2 3 The tension graph The upper graph is the tension graph that normally shows two lines The bright blue line indicates the tensions recorded by the upper tensiometer ver
43. capillary part is here expressed as 9 0P 6 Thus 0 resembles the classic soil water retention function with residual water content 1 1 1 Adsorptive saturation function The saturation function for water adsorption is given by a smoothed piecewise linear function Iden and Durner 2014 s x 1 fx x bIn f exp 2 Xa Xo where x pF log Ch in cm Xa log o ha and xo log No with ha being the suction at air entry for the adsorptive retention and ho being the suction where the water content is 0 The parameter ho might be set to 10 8 cm which is the suction at oven dryness for 105 C Schneider and 50 HYPROP FIT Software Goss 2012 The smoothing parameter b is dependent on the capillary saturation function S which is described next 1 1 2 Capillary saturation function The formulation of the saturation function for capillary saturation SP must ensure that the water content is O at h ho This is guaranteed by scaling the basic classic saturation functions I h e g van Genuchten 1980 Kosugi 1996 SAP h Aio 3 where Ip is the basic function at h ho Setting Io O yields the original capillary saturation functions where S h T h 4 Thus SHYPFIT2 0 distinguishes between scaled equation 3 and unscaled equation 4 capillary retention The effect of scaling according to equation 3 is exemplarily shown for the van Genuchten function in Fig 2
44. cssadaseteieatdecactentouiaenidecsentouseetas 53 12 FIr EOndUCTIV IY eeen E E E A AO 55 1 2 4 Prediction of isothermal vapor conductivity seseseseseeseseeesssreresrrersrrrresrreresrereserereserresene 56 TCD C UCO a E 57 2 Other soil hydraulic TUNCUOIS sorsrreepirrne n aS 59 2 1 Brooks and Corey UNCON esar nne soa huncioasetnctorsiunesaosuenesteeneee 59 FD POC UC IC A Ne OOS e 59 3 Fel SSI a Oa cnsse gee raster toc eceacnsesetecienngeecacwsaecasn ssa eacnsaeacnnsaaeerasetet Eaa REES 62 5 DET IOMHOT ODjCCUIVE TUNIC UO epii E E 62 3 4 Integral VERSUS classic BIE sxccecsssmcaccucsncasesscndcoadseneasascencasacseedasas psedseuiasnsaeencsecsmncsaeienctsandseiissneasendeeutes 62 JaN LENNE eld 1 oc eee oe ene enn eee ee ee ee 63 BPOSC Dl OCC SSIS saiucussbadcorucncncdcineaseuraveaveusiotaiestanesncleunntdestunbinntonmbincotisbunn dene E i 65 ALDENO CVN DE eee ene eee ene A eee eee oe 65 Ar WIC OU Cal NEY ete tI S ee se cercaeeressint tiered uo nesses ta roe we rrsetvaneeG oases veo ude se ea aree a 66 4 2 1 Parameter uncertainties es sssensseesesresseesssrsssresseosssrsesrosseosssesssresseosssesesrosseosssesssresseess 66 4 2 2 Function uncertainties ssssesssessssesssresseessersssresseosssesssrosseosssesssrosseosssesssresseesseesssresseese 66 RETEN O E A EA E E E E AE EE NE 67 HYPROP FIT Software Da HYPROP FIT Software 1 O0verview and Scientific Background HYPROP FIT is a software tool for Windows environme
45. d from the graphs see section 4 4 4 Fit quality is given in terms of the root mean squares errors separately calculated for the water content data RMSE_TH and the log of conductivities RMSE_K The root mean squares error is an indication of the mean distance between a data point and the fitted function An RMSE of 0 001 for the retention data fit indicates an average distance of the fitted curve to observed data of 0 1 water content Models with a higher number of parameters are generally more flexible hence the fitting error will become smaller for models with a larger number of parameters This is counterbalanced by a higher correlation between the optimized parameters and higher uncertainties for the individual parameters As an aid in deciding which model choice is most appropriate HYPROP FIT shows in the Statistical analysis field additionally the corrected Akaike Information Criterion AlCc Akaike 1974 This criterion accounts for the different number of adjustable parameters when selecting the best model The value will be normally negative The smaller the value i e the larger the absolute number the more appropriate is the model For details on the Statistical criteria see Peters and Durner 2008 Parameter The effect of the parameter uncertainties on the uncertainties of the fitted functions can furthermore be visualized in the graphs by a grey shading of the uncertainty bands This 35 HYPROP FIT Software option
46. d function Therefore the derivatives are given for the unscaled functions The derivatives are dh for the unimodal van Genuchten function ar ex dh V2T0h J 1 el a anm ah t1 ah ort D dr n _ e 20 dh V2T0h for the unimodal Kosugi function and simply avy ar dh t 1 t an 57 39 40 41 42 HYPROP FIT Software for the bimodal functions 58 HYPROP FIT Software 2 Other soil hydraulic functions The above summarized hydraulic functions encompass most of the commonly used soil hydraulic functions such as the original van Genuchten Kosugi and Durner functions and extend them to account for water retention and conductivity in incompletely filled pores as well as for isothermal vapor conductivity Sometimes other models which are not part of the PDI model family are required such as the Brooks and Corey 1964 model or the model from Fredlund and Xing 1994 These two models are outlined here 2 1 Brooks and Corey function The Brooks and Corey retention function is given by ah for h gt a Seth e for h lt a a where a 1 cm is the inverse of the air entry value A is the pore size distribution index and Se is the effective saturation given by S 0 0 0 0 with 0 and 0 are saturated and residual water contents Based on similar capillary bundle considerations as the Mualem model the combination of the Brooks amp Corey retention func
47. d in cm per day Also please care for the regional conventions for the decimal separator comma or dot as specified in the Windows System Software Added data can be edited and deleted Upon a right hand mouse click a pop up menu appears Delete row Delete content Add Points From File As an additional feature retention and or conductivity points can be read from data files of type csv This is of particular advantage if many points are available from independent measurements 31 HYPROP FIT Software 4 3 4 Modifying evaluation parameters Power User only Edit Evaluation Parameters O Control parameter for sensitivity analysis Interpol hennitian_splines Number_of RETC_data 100 Pohnomial_degree 5 Tensiometer_Mean geometric Tensiometer_Mean Detemines the calculation mode for the mean tension for the calculation of the water retention curve points default geometric Canes The parameters for raw data interpolation can be edited in a submenu i which pops up when clicking on the respective button on the right side of the header line in register Evaluation The interpolation parameters are described below General Information Variable Explanation possible values Interpol Determines the interpolation mode for the measured data 0 polynomial 1 piecewise linear 2 hermitian splines Number _of_ Number of support points for hydraulic RETC_data functions lt 1000 Polynomial degree If
48. default all parameters are allowed to vary 4 4 3 Results of the parameter estimation and uncertainty measures Soil hydraulic model selection Retention function parameters Conductivity function parameters ae bimodal Parameter Value 2 5 97 5 Unit Parameter Value 2 5 97 5 Unit original PDI bimodal PDI aipha 0 00729 0 0069 0 0077 1 cm Ks 0 534 0326 0 875 cm d Brooks Corey in 2 455 2 282 2 651 tau 1 599 1 846 1 351 Fredlund Xing C C Cc t th_r 0 119 0 109 0 129 cm cm pn pn p v ith_s 0 338 0 333 0 342 cm cm Kosugi van Genuchten m 1 1 n f C C C D AAMAAAAAAAAAAAAAMAAAAAA van Genuchten mnvar C Model VG Model Code 1100 Source van Genuchten 1980 Description traditional constrained van Genuchten Mualem model After termination of the fitting process the appearance of the parameter fields switches from the input mode to the results mode In results mode the optimized parameters are shown in the previously empty column with the Value fields and the uncertainties of the individual parameters are indicated by 95 confidence limits for the parameter values expressed by the 2 5 quantiles and the 97 5 quantiles in the two following columns For fixed parameters the uncertainty is of course zero Statistical analysis RMSE_TH 0 0046 RMSE_K 0 1094 Me The quality of the fits is quantified in the group Statistical analysis in the upper right and can be visually inspecte
49. e tvp file Notes Notes _ Description This data group contains free text fields On the left side some pre defined entries are listed which can be useful for a later classification possible use of data It is recommende to fill them out since they are uniquely related to the measurements data set On the right side free text can be inserted and edited that is informative for the respective Site description user and its purpose General data Site coordinates 22 HYPROP FIT Software 4 2 Register Measurements Visualization and editing of measured data File Extra Help 0 G1 infomation 2 Measurements A Evaluation Ce Fitting Q Epot Search Stop Point Use Air Entry Point Search Air Entry Point Tension bottom Tension top T a c 2 pa 5 Sat 10 Sun 11 Mon 12 Tue 13 Wed 14 Thu 15 Fri 16 Jan 2015 Time Weight Weight g 5 3 iS gt Show Spline Interpolation Points Show Data Points Date Time 09 01 2015 10 58 17 09 01 2015 10 59 17 09 01 2015 11 00 17 09 01 2015 11 01 17 09 01 2015 11 05 02 09 01 2015 11 06 02 09 01 2015 11 07 02 09 01 2015 11 08 02 09 01 2015 11 09 02 09 01 2015 11 10 02 09 01 2015 11 11 02 09 01 2015 11 12 02 09 01 2015 11 13 02 09 01 2015 11 14 02 09 01 2015 11 15 02 Wed 21 09 01 2015 11 16 02 Sat 17 Sun 18 Mon 19 Tue 20 Date Time 09 01 2015 11 07 39 09 01 2015 12 02 20 09 01 2015 13 32 33 10 01 2015 09 49 49 11 01 201
50. e asked for A new directory can be created and specified by pressing the respective button Depending on the measurement mode Single balance mode one balance for more sensor units or Multi balance mode one balance per sensor unit one or more bhdx files with the names of the measurement campaigns will be created in the target directory 3 2 2 Menu File Opening modifying and storing HYPROP projects The menu File is used to open and save project files and to import measurement of retention and conductivity data An existing data file is loaded by selecting the respective file name in the open window as described below After opening a file five registers will appear The registers are entitled Information Measurements Evaluation Fitting and Export Save Ctrl 5 ohd Import bhdx bhdi bhdix CSV bhdi bhdix Open is the uppermost entry in the menu File Four basic types of data files can be opened HYPROP FIT binary project file containing information measurement data calculated conductivity and retention data and fitted functions for a single measurement campaign DEFAULT A HYPROP FIT iteration project file including data and fitted functions for multiple measurement campaigns csv files that contain HYPROP FIT data in a structured manner These files can be created in the Export register of the software see section 5 external
51. ective buttons that are located immediately above the graphs to the left Switching a graph on or off will resize the remaining graphs 4 3 1 Calculation of retention and conductivity data The calculation of retention and conductivity data from the measured tensions and weight changes follows the simplified evaporation method as outlined by Schindler 1980 Schindler et al 2010a b Peters and Durner 2008 and Peters et al 2015 For the theoretical background of the calculations the reader is referred to this literature In the calculations only the tensions and weights in the time window between start time and stop time which is specified in the previous register Measurements are considered 4 3 2 Calculation of absolute water contents The relative shape of the water retention data is fully determined by the recorded weight change The absolute level of the water retention data is additionally affected by the parameters that can be specified in the parameter group Calculation of water contents 29 HYPROP FIT Software Option1 Calculation of the initial water content from dry soil weight Calculation of water contents G From dry soil weight g 410 0 From given initial water content vol Estimation of initial water content Calculated Parameters Initial water content Vole 37 3 Dry zoil weight g 10 0 Density gems 1 65 Porosity 0 38 This is the option of choice if the dry soil weight
52. eeeeeeeeeeees 35 AAA VEY PVCU Mi INS a cb tence etree cca tenses E acne aoeeea 36 SI HYPROP FIT Software 4 4 5 Field capacity and plant available water cccccccssccccsssecccesececeenecceeeeseceeeeeceeeeeceeseneeeeees 37 4 4 6 Modifying optimization parameters Power User ON y ccccccssssecceeeeessecceeseeeeeeeeeeaeeseeess 37 4 5 Register Export Exporting data graphs functions and PArAMELELSS cccccessececeeeeeeeeeees 39 we POSC ne LBL 8 Ors irs Jc ES E ee eee 42 O PROM CCS a ns cacti wttina te cations aaa E T E ena E E 43 APPENDIX 1 Theoretical basics of the Simplified Evaporation Method cccccccssseceeeseeeeeeeeeeeees 44 APPENDIX 2 Automatic estimation of the initial water content ccccccccccssssesseeececeessaeeseeeeceeeenaaas 48 APPENDIX 3 SHYPEITZ 0 USEF S IVIGNUGN errena 49 1 POI model GonDINatIONS sports perc ea pesca steyen sa ece cn dseucnsss T aan oheqeerasevncsanguareneasreeninees 50 LIRE enuon anc secssetecasteecuttescncteletevaukscanescsisbesd venreeesutt E E 50 ON E E ET alls E EE E E E A E A A E E 50 1 1 1Adsorptive sat ration TUNCON sercan EA 50 1 1 2 Capillary saturation function eessesssseseseesesrrrsssreresrresssreressreessreressrensserressreresreressrereserresere 51 L2 Conductivity UCONN eie T E TE 53 DD oN SUVS a E EAE AANE E EEEE AA E A EEEE ET EN A E 53 1 2 2 Capillary conduCtVily ced tictdacsapseserssenbdesuentesensteahduconsdeainndeueete
53. es for most specifications Power Users are experienced and skilled users trained in soil physics and familiar with the scientific background of the data processing A change of the user mode is done by selecting the appropriate mode at the lower right corner of the screen in a drop down menu A change of the user s status can be done at any time and from any register It will immediately be active Power Users are allowed to alter many of the default values or to manipulate measurement data As an example power users can edit or delete measured data records that are obvious outliers or delete undesired records at the beginning or end of a measurement campaign Furthermore power users can process data that are measured on different devices e g for evaporation experiments with different soil column heights Changed parameters will affect the results of the data processing and the implications on the calculations must be known by the users In other words if a Power User decides to alter parameters he or she should really know what they are doing SI HYPROP FIT Software 3 Starting the Software and Loading a Project 3 1 Starting the program To start the software double click the respective icon or launch the software from the menu of installed programs Upon the first start an empty welcome screen will appear AJ HYPROP Data Evaluation Software File Extras Help www ums muc de IO File explorer Gy 2 Power
54. f 2 o9 2 i ev f erfc7 war 4 k Miimi i D I l Kre Xi 1 Wi spar T 16 2 Die Wine for the bimodal Kosugi function 1 2 3 2 Analytical solutions for scaled capillary retention functions For the scaled capillary retention functions the Mualem model can be written as Peters 2014 Ko S SP fronti 17 Ire zar where and I is a dummy variable of integration This leads to the following general solution 2 cap cap capt oe Ke S S Fossa 18 where F T is the solution of the indefinite integral fod F T for the unimodal constrained van Genuchten and for the unimodal Kosugi function are given by _ r1i m se F T a 1 r 19 and FO 2 ert lerte ar 2 20 r S erflerfc 4 20 where erfc 1 is the inverse of the complementary error function For the bimodal van Genuchten and Kosugi functions Durner 1994 and Romano et al 2011 F T is given by F P Y2 wa 1 ri 21 and FT 2L w erf erf 2r P 22 54 HYPROP FIT Software Equations 18 to 22 yield the following solutions for the capillary conductivity function 2 1 m Kear sey 1 SE 23 lo for the basic van Genuchten function 2 Fo erf erfe 2 Z cap _ cap t V2 with Oo Fy erf ferf 21 Z V2 for the basic Kosugi function m l _ xe wia 1 p Kre Si 1 Wi om 1 25 a wia 1 T i i for the bimodal van Genuchten function and Fo y Wig evi
55. f water contents at 6 and at 1500 kPa PAW 6 1500 kPa plant available water as difference of water content at 33 and at 1500 kPa 4 4 6 Modifying optimization parameters Power User only Edit Fitting Parameters PFBOUNS_LOWER PFBOUNS_UPPER Show_Uncertainty_Bands WEIGHT_FLAG wk wtheta Number_HYDFUNC Para meter Number _HYDFUNC PFBOUNS_LOWER PFBOUNS_UPPER E Control parameters for multiobjective estimation 500 described below Number of supporting data points to depict the hydraulic functions default 500 Defaut Cancel Description The hydraulic functions in the graphs are taken from internal tables where the functions are listed in the range PPBOUNS_ LOWER to PFBOUNS_ UPPER at discrete supporting points Number_HYDFUNC is the number of supporting data points to depict the hydraulic functions Not This affects also the number of data written in the exported tables see section 4 5 Lower bound for internal table of hydraulic functions expressed as pF logio h with h in cm The default value expresses a minimum tension of 0 001 cm Upper bound for internal table of hydraulic functions expressed as pF logio h with h in cm The default value is pF 7 which is a tension of 10 000 000 cm 37 Power Users can edit some of the parameters that specify the calculations and the depiction of the fitted functions The specifications are listed in the menu Optimization
56. functions P A Yi wil 8 where I are the weighted subfunctions of the system expressed by one of the unimodal functions and wi are the weighting factors for the subfunctions subject to 0 lt wi lt 1 and w 1 Thus 3 basic functions which are either uni or bimodal are implemented These functions can be scaled according to equation 3 or not scaled Furthermore the adsorption function might be added or not in that case S in equation 1 is simply set to 1 Allin all this gives 24 combinations for the soil water retention function basic functions 3 modality 2 scal 2 ads 2 are possible See Tab 1 The unscaled unimodal van Genuchten function without adsorption first line in Tab 1 is the original van Genuchten function The shape parameter b for the adsorption function equation 1 is given by b 0 1 nae exp N 9 for the constrained and unconstrained van Genuchten functions and by b 0 1 0 070 1 exp N 10 for the Kosugi function For the bimodal functions n and o are taken from the coarsest subfunction i e the subfunction with highest value for a or the lowest value for hm Figure 3 shows a bimodal scaled water retention functions with capillary and adsorptive part 52 HYPROP FIT Software 5 10 10 10 10 10 10 10 h cm Figure 3 Typical bimodal PDI soil water retention function The capillary part is here expressed as 9 0P 6 Thus 0
57. ght of the soil is not yet available the initial water content of the campaign can be estimated automatically by the software assuming an initially full saturation of the pore space i e 30 HYPROP FIT Software initial saturation 100 In practice this will rarely be reached the calculated water content is therefore to be seen as an approximation of the true one As a result of the assumptions the initial water content the identical porosity the dry bulk density and the dry soil weight can be calculated These values are shown in the underlying group Calculated Parameters Details of the automatic calculation are given in Appendix 2 If the mode of the initial water content calculation is changed the user edited values for the dry soil weight and for initial water content remain in storage This allows a user to switch between the options and to conveniently compare the outcomes of the different calculation options without losing the inserted information IMPORTANT The calculation of porosity and absolute water contents depends on the correctness of all involved parameters that are specified in the Information register in particular on the correct dry bulk density DEFAULT 2 65 gc m mass of the measuring head DEFAULT 358 g mass of the steel cylinder that contains the sample DEFAULT 200 g and mass of dry soil which is specified in the Evaluation register Any error in these data will be linearly transferred
58. h reflects the evaporation rate in the laboratory environment Small changes in the slope can be caused by temperature fluctuations in a lab with the respective changes in relative humidity The period of a constant weight loss rate is called stage 1 evaporation and water loss during that phase is controlled by atmospheric conditions Afterwards the rate drops and the slope of the weight curve becomes less steep This second phase is called stage 2 evaporation The water loss rate is now controlled by the resistance of the drying soil near the soil surface In the depicted data example the stage 2 evaporation starts at the time when the upper tensiometer reaches its air entry value For further information see Peters and Durner 2008 and Schindler et al 2010 2011 27 HYPROP FIT Software 4 2 6 The tension data window The window at the upper right shows the recorded tensiometric data together with the times and the temperature The latter is for information purposes and not used in the data evaluation The data window can be scrolled by the scroll bar to the right By moving the cursor into the data window and clicking on the right hand mouse button the data can be directly edited by Power Users Possible operations are deletion of lines insertion of lines and changes of the contents of individual data cells Also multiple lines can be selected and the data can be copied with the ctr c option Manipulation of data or
59. his is useful to compare different measurements and to fit retention and conductivity functions to multiple sets of data Individual data sets will be shown in different colors and can be switched on and off HYPROP FIT Software 3 2 4 Menu Help HYPROP FIT version information The Help Menue consist of three entries User s Manual This entry gives direct access to the pdf version of this manual This entry checks for updates of the HYPROP FIT software It requires Internet access If an updated version is available it opens a download wizard that helps to get an installation version of the updated software Check for updates The Info lists the HYPROP development team and shows the disclaimer 14 HYPROP FIT Software 4 Processing Data The HYPROP FIT software groups the data processing in up to five registers that encompass all steps of the data evaluation in a logical sequence of consecutive operations Normally the registers are followed consecutively from left to right In all registers data processing options can be specified and data can be edited or manipulated Information ma Measurements AL Evaluation a Common to all registers are the following points Editable values are shown in a edit field i e surrounded by a frame with a slightly brighter background Values that cannot be modified are shown without the edit field Whether or not a variable can be edited de
60. ing Filename fomat lt Date gt _ lt Samplename gt lt Samplename gt lt Date gt Directory C Users wdumer Dropbox UMS_Software data tests OFFSET Tests Serie 6 Chats Data Retention Conductivity Kitheta Spline Points 18 32 03 Tension Results bamm 25 17 93 42 J Information Measured values pF EJ Water content Vol pF log 10 K fem d Water content Vol log 10 K em d KIKIKI The register Export allows the export of all graphs raw data calculated data fitted functions and other parameters of interest User specifications for data format filename convention and directory will be kept as default as long as the user remains in the session Export Format The format of the exported data is defined in the box on the upper right HYPROP FIT supports the following output formats CSV comma separated values for editing with any external program csv csv simple comma separated values for re import into a HYPROP FIT software Creates three files sdexe lt name gt _Config csv lt name gt _Tension csv HYDRUS Table lt name gt _Weight csv which contain the respective data These files could be manipulated outside of HYPROP and re imported through the File Import menue xls or xlsx Microsoft Excel file format The Excel file will contain all relevant data in different sheets e Configuration and information e Measured raw data e Data from the spline i
61. is only available for Power Users by selection the respective option in the Optimization Parameters menu see section 4 4 6 Parameter eee ee As indicated above models with a higher number of parameters are generally more flexible But will have a higher correlation between the optimized parameters and higher uncertainties for the individual parameters Pressing the button Parameter Correlation Matrix will cause the pop up of the parameter correlation matrix Correlation Matrie z alpha n alpha 1 000 0 645 0 649 1 000 thor 0 339 0 895 th_s 0 836 4 446 tau 0 017 0 014 Ks 0 332 4 234 In this symmetric matrix the linear correlation coefficients of the parameters are listed Negative values indicate that the increase of one parameter value can be partly balanced by a decrease of the other positive values mean that an increase of either parameter value has a similar effect on the overall fit of the functions Generally correlation coefficients with absolute values smaller than 0 95 are not problematic and indicate a well posed fitting problem Too high correlation of parameters is undesirable since it indicates an over parameterization of the selected model 4 4 4 The function graphs In the lower part of the Fitting register window graphs with the calculated HYPROP data the added data and the fitted hydraulic functions are shown The individual graphs can be switched on or off by pressing the respective butt
62. ly modified and re imported A HYPROP FIT iteration project file including data and fitted functions for multiple measurement campaigns 11 HYPROP FIT Software File Extras Help gt Open Ctrl 0 Fl Save Ctri S Import File Extra Help Ctrl 0O Save Ctri S File Extras Help ce o pen ipl Save Ctrl 0 Ctrl 5 File E Open Ctrl O Save Ctrl 5 Ey Save as 1 CX KI bhdi 2 Ci GG bhdin Yi Fe l Clase Import All files Shows files with any extension Only files with the selected extension will be shown in a browser window and can be opened by double clicking on the name Save is used to save a file which has been imported or processed All changes will be saved If no changes to an existing project have been made the save button is inactive grey shaded Save as is used to save a file with a different name Please note that you can open and process a file already during the stage of data acquisition with HYPROP VIEW However this Original file is write protected as long as the measurement campaign continues If desired you can save it with a different name iett 1 C Gruppell bhd Import Comma Separated Value csv Import Data for Fitting Only Import is used to import data from csv files of type config csv which have been previously extracted and edited outside the HYPROP software and are now re imported
63. n bottom Tensiontop Air Entry Point botom Air Entry Point top Reduced Data Points bottom Reduced Data Points top i Air Enjry Point to ml 3 Tension hPa at top top Stop bottom Sat 10 Sun 11 Mon 12 Tue 13 Wed 14 Thu 15 Fri 16 Sat 17 Sun 13 Mon 19 Tue 20 Wed 21 Jan 2015 Time 4 2 4 Using the air entry point of the tensiometer cup as additional tension measurement Power User only It is possible to extend the range of the tensiometric measurements by using the air entry value of the ceramic tensiometer cup as additional measuring point Schindler et al 2010b The HYPROP FIT includes this option for Power Users The air entry point is located at the time where the sudden pressure drop of the tensions to zero occurs Figure Of course this option is only viable if the sharp drop towards zero tension is really recorded i e if the measurement campaign was performed long enough Clicking on the Use Air Entry Point button will add for each tensiometer that has reached the point where air passes the tensiometer cup an extra measuring point At that point in time a tension that is specified in the register Information HYPROP Parameters is used as measure tension of the respective tensiometer The button Use Air Entry point is an on off switch Hence to deactivate the use of the air entry point press the button again At the time axis of the air entries of the two
64. ng process and allowed to vary within a predefined range The ranges of values are wide enough to cover physically consistent parameter combinations A user can alter the permissible values of the bandwidths by changing the default minimum and maximum value but for most soils this will not be necessary In cases users may have prior knowledge about parameters that can lead to much reduced bandwidths which speeds up the convergence of the algorithm Changing the default values for the minimum and maximum parameter values is done by typing the new value into the respective field in the group Parameters retention function and Parameters conductivity function To store a typed value the cursor must be positioned out of the field which can be done by pressing the Tab or Return key or by moving the mouse to another field If a hydraulic parameter should be kept at a pre defined value this value can be inserted in the respective field Automatically the tick box in front of the respective parameter will be de selected 1 In cases where the fitting will apparently not lead to a stop the user can cancel the process from the Windows Task Manager 34 HYPROP FIT Software NOTE For the decimal separator i e appropriate use of comma or dot in the inserted numbers the regional conventions of the Windows system is used If a model selection is changed all specified parameter values will be automatically reset i e by
65. nown from typical Windows GUls To perform the actions position the cursor on a command or entry The selected entry will be highlighted It is launched by a single click with the left mouse key Some commands are on off buttons An example is the File explorer button in the third header line Clicking on it will lead to the pop up of a file explorer window on the left of the main window clicking on it again will lead to its disappearance Other commands open submenus where further commands or options can be selected Commands that cannot be actively selected in a certain situation are shown in grey and cannot be used As an example the Save button can only be used if a project has been opened and modified Draw down menus for selecting pre defined options are indicated by vertical triangles on the right side of the respective button Entries can sometimes be selected by key combinations If so the combinations are indicated to the right of the respective entry Menus can be left without any action by pressing the Escape button Three types of data can be processed e Raw data from TensioView measurement campaigns tvp performed with the HYPROP experimental setup e Existing projects of type HYPROP binary project bhd or bhdx HYPROP binary iteration project bhdi or bhdix or ASCII files _config csv with a predefined data structure e Separately measured retention and conductivity data in ASCII files _RETC csv
66. nt The software reads data that are recorded in projects by the data acquisition software HYPROP View and stored in lt tvp gt files The files are converted to HYPROP FIT files with the extension lt bhdx gt single measurement Campaigns or lt bhdix gt multiple measurement campaigns Furthermore separate data sets consisting of measured retention data and conductivity data can be imported and fitted creating project files with the extension lt fitx gt HYPROP FIT performs the following operations i Specification of all required parameters for the evaluation of the recorded experimental data with the simplified evaporation method SEM such as column length positions of tensiometers tare weights of the measurement device components and so on ii Visualization of the measured raw data i e tensions and weight changes dand specification of starting and stop points for the data evaluation iii Re calculation of tensions and net weight data as base for the calculation of retention and conductivity data This includes the temporal interpolation for data in low temporal resolution and the aggregation of data in very high temporal resolution iv Calculation and visualization of the data for the retention characteristic and the conductivity characteristic v Fitting of state of the art hydraulic functions to the data visualization of the functions and listing of the values and confidence limits of the hydraulic par
67. ntents Add retention data points Add conductivity data points Add WP4 data points Edit Interpolation From dry soil weight g 387 5 on a r Parameters From given initial water content Vol FE ak Weighting ep log10 K Weighting z ea Gross dry 13 Weighting ull W Q Estimation of initial water content a Vol factor oF ES em d factor PFH iol mass mass o factor Maset Vv ol Calculated Parameters Initial water content Vol 34 9 Dry soil weight o 387 5 Density g cm3 1 56 Porosity 0 41 Show data as small points Retention O pF Conductivity K pF Conductivity K O data added data z data added data data log 10 K in cmid log 10 K in cmd vol water content 4 0 5 10 15 20 25 30 35 pF vol water content PowerUser The register Evaluation shows in the upper part four data groups with specifications for the calculation of absolute water contents calculated parameters and three menus for adding additional data points In the lower part up to three graphs of the resulting retention and conductivity data are shown The first graph shows the volumetric water content versus pF decimal log of tension expressed as pressure head in the unit of cm The second graph shows the log of the hydraulic conductivity versus pF and the third graph the log of hydraulic conductivity versus volumetric water content The individual graphs can be switched on or off by pressing the resp
68. ntepolatio nused to derive retention and conductivity data e Retention and conductivity data e Data of fitted curves e Parameters of fitted models e Uncertainties of fitted model parameters e Base points if specified 39 Ss HYPROP FIT Software HYDRUS Table File namend MATER IN readily formatted for direct use in HYDRUS simulations The file contains retention curve specific water capacity and conductivity data Different output settings are available and described below The exporting is executed by clicking the button Export Filename format Enter the output file name in the field Filename format All permitted characters can be used in the filename The two default settings lt Date gt and lt Samplename gt are available for optional use in any position within the filename They will automatically be replaced by the current date resp sample name Directory Enter the output directory in the field Directory or click on Browse to select a directory Content Content v Charts J Data J Retention v Information 4 Base points v Conductivity v Measured values pF Water content Vol pF log 10 K cm d Water content Vol log 10 K lem d v Kitheta v Results v Tension v Weight __ Update In the group Content you can select check boxes for the data that should be exported and enter the base points for export In the subgroup Chats you can select check b
69. oil sample Default value depends on 5 0 cm No Yes height selection of cylinder type Underfilling of cylinders or swelling of samples can require a modification of this entry Position Height level of the measuring tip of the tensiometer 1 25cm No Yes lower with the short shaft measured from the base of the tensiometer soil sample Default value is valid for the UMS HYPROP device where the lower tensiometer is located at of the total height Position Position of the upper tensiometer in the sample 3 75cm No Yes lower measured from the base of the soil sample Default tensiometer value for the UMS HYPROP device is of the sample height Soil volume Volume of soil sample It is calculated from the entries 249cm No No of area and height and corrected for the volume of the two tensiometer candles 1 cm 18 HYPROP FIT Software HYPROP Parameters HYPROP Parameters Empty soil sampling ring weight g 201 7 Measurement head net weight a 363 9 Air entry pressure upper tensiometer bar 9 9 Air entry pressure lower tensiometer bar 2 9 Density of solid substance g cm3 3 65 Empty soil sampling ring weight Measurement head net weight Air entry pressure upper tensiometer Air entry pressure lower tensiometer Density of solid substance Exact mass of the soil sampling ring dry and clean IMPORTANT THIS VALUE MUST BE SPECIFIED INDIVIDUALLY FOR EACH MEASUREMENT The mass can be determined bef
70. olumn center 2 For time t _ nee 2 be half the evaporation rate doi L i i l Se cick gg a a aa i where At _1 t ti and z runs from 2 to N the number of data points The 22 2 mean gradient of the hydraulic head at the center of the column is then approxi mated as _ 2 OPE z Ng i 1 2 lt 3 Z1 where hy jy 1 and hg i 1 are the mean suctions at the two depths z and z3 in the time interval At _ 1 Which are calculated by the arithmetic mean in time The hydraulic conductivity K emd is calculated from the Darcy Buckingham law as lt 1 Q gt N e K 2 3 1 2 These conductivity data are assigned to mean suctions h to obtain point data of K h If the arithmetic mean is chosen to express the effective mean suction h is calculated as the arithmetic mean of the four suctions used for the calculation of the hydraulic gradient Peters and Durner 2008 Schindler et al 2010a 4 5 1 t 3 9 Hence point data of the conductivity function are given by K _1 h _1 For the data of the retention function it is straightforward to assign the temporal arithmetic 1 This provides triplets of 0 2 K and h which leads to the full set of 0 h K h and K data mean of the two water contents 6 _ and 6 to hari i 46 HYPROP FIT Software A limitation of the evaporation method is that at the beginning of the ex
71. ons that are immediately above the graphs to the left Switching off a graph will resize the remaining graphs As with any other graphs in the HYPROP FIT software the graphs can be magnified to full screen or zoomed by selecting the respective option in the context sensitive submenu which is activated by a right mouse click The fitted functions will disappear from screen if either parameter settings are changed e g by modifying the permissible data range or a setting in another register is changed that affects the calculated HYPROP data After the fitting process the fitting results can be exported in a variety of ways and graphs which is specified in the final register Export 36 HYPROP FIT Software 4 4 5 Field capacity and plant available water Field capacity WC 6kKPa 32 2 WC 33kPa 176 WC 1300 kPa 11 9 Retention curves are often determined to get estimates of field capacity PAW 6 1500kPa 20 3 PAW 33 1500 kPa 5 7 and encom passes and plant available water HYPROP FIT shows immediately after each fit the corresponding values The calculation is based on the fitted curves WC 6 kPa water content at pF 1 8 i e field capacity for conditions with close groundwater level WC 33 kPa water content at pF 2 5 i e field capacity for conditions distant to groundwater WC 1500 kPa water content at pF 4 2 i e at the wilting point PAW 6 1500 kPa plant available water as difference o
72. ooks and Corey Fredlung Xing Kosugi and van Genuchten with and without parameter constraint f r the parameter m With the exception of the Brooks Corey model all functions are available as PDl variant in uni and bimodal form The PDl variant ensures that the water content matches zero at oven dryness and it considers the effect of water in capillaries in films and in corner of pores in both the retention function and conductivity function All models are described in detail in appendix 3 4 4 1 Fitting functions W Fitting The Fitting register contains a header line with the fit button at right Curve Fitting and a button for editing of Optimization Parameters The data to be fitted are shown together with the fitted functions in the graphs in the lower part of the window The non linear parameter optimization is subject to constraints and settings which are defined in the submenu Optimization Parameters see 4 4 3 To fit the selected hydraulic functions see 4 4 2 to the data click on the Curve Fitting button A window will pop up that indicates that the fitting routine is running This process can take seconds for 33 HYPROP FIT Software the simple models with few parameters up to minutes for complex models with many parameters and cannot be stopped or otherwise interrupted Fitting Evaluation is camied out This process can last some minutes and can not be cancelled PL The fi
73. ore or after the HYPROP measurement campaign The default value is the mean value of the first series of sample rings provided by UMS Wrong values transfer linearly to wrong calculated total water contents An error of 2 5 g leads to a shift of calculated water contents of 1 Mass of the dry clean measurement base The mass includes the two filled tensiometers and the silicon sheet that is used to separate the head from the measurement base but with no water below the sheet Normally the weight is accurately derived from the default that is dependent on the serial number of the head If uncertain then perform a control measurement For HYPROP devices that are long in use it is advised to weigh them separately Wrong values transfer linearly to wrong calculated total water contents An error of 2 5 g leads to a shift of calculated water contents of 1 This value specifies the air entry pressure of the cup of the upper tensiometer The value is used if the use air entry option is activated power users only The default value is the statistical mean of ceramic cups used in the UMS HYPROP tensiometers For individual specifications please follow the detailed instructions given by Schindler et al 2010b As above for tensiometer at lower position The density of the solid soil material is by default set to the density of quartz i e 2 65 g cm3 which in general is a good approximation for a wide range of soil materials
74. oxes for the diagrams that should be exported In the subgroup Data you can set the options for exporting raw data calculated values and fitted functions Furthermore you can freely define W Ease points Base points are specified pF values for which water contents will be calculated An example is the calculation at field capacity of witling point To define base points enter the desired pF value in the table and the corresponding water contents will be calculated when the butten at the lower right is used Base point data will be included in the export You can store the base point definition in a template by selecting a template name and directory in the upper line 40 HYPROP FIT Software Post Processing Cin mothinc Do nothing Show in Explorer Open main exported file Go to Export Folder Now In the group Post Processing you can set different actions that should be executed either after the exporting or immediately at exporting History History umsserver1 user thomas Documents HYPROP 110420Demo1 csv created umsserver1 user thomas Documents HYPROP 110420Demo1_Retention png created umsserver1 user thomas Documents HYPROP 110420Demo1_Conductivity png created Numsserver1 user thomas Documents HYPROP 110420Demo1_Ktheta png created umsserver user thomas Documents HYPROP 110420Demo1_Tension png created umsserver1 user thomas Documents HYPROP 110420Demo1_Weight
75. pendent on water content and is calculated according to Saito et al 2006 where is the volumetric air content Da m s is the diffusivity of water vapor in air and is the tortuosity factor for gas transport calculated according to Millington and Quirk 1961 gi ary 32 D and psx are dependent on temperature T 2 D 2 14 10 33 273 15 and p 10 exp 31 3716 S187 7 92495 10 T r 34 H is calculated with the Kelvin equation hMg H exp 35 i o a 35 Figure 4 shows total hydraulic conductivity and the single parts for an unimodal and bimodal capillary retention function 56 HYPROP FIT Software 10 10 os o E 10 10 5 o lt lt 10 10 9 Figure 4 Total hydraulic conductivity for unimodal left and bimodal right capillary retention function 1 3 Capacity functions The soil water capacity function is defined as d C h 7 The general form of the PDI capacity function is dscoP dsad C h 6 E 0 r dh P Ev l _wewwes Adibi AAL r rerewe i 1 10 10 10 10 10 10 10 h cm where the derivative of the adsorptive saturation is given by asee _ 1 dh hIn 10 xa Xxo _ 9s Or at ase 1 To dh r dh exp 4 1 exp 4 36 37 38 The difference between the derivatives of the scaled and unscaled capillary saturation functions is only the denominator 1 Ig which is 1 for the unscale
76. pends on the user mode power users can edit more fields than public users All variables that are expressed by numerical values are given with units If a value is edited by the user the inserted numerical value must be in the given unit When inserting a numerical value the proper decimal separator sign depends on the region settings of the computer Default for the German Operating System is the comma whereas default for the English one is a dot Fields for numerical values do not accept input of alphabetical letters or special signs if a user attempts to insert a non permitted character the field does not show the typed character To be sure that an input operation is finished the cursor must have left the input field This is achieved either by pressing the Tab Key or the Return key or by moving the curser to another position in the window and clicking the left mouse button In some menus explicit pressing of an Accept button is required When placing the cursor on a button a specific text information Quickinfo about the button s function pops up for a period of some seconds Some of the graphs and other contents have context sensitive menus that are activated by placing the cursor on them and a right click of the mouse 15 HYPROP FIT Software 4 1 Register Information Specifying the required parameters for the measurement File Extra Help TO o E riomston FZ Measreners S Evatuaton Fo Peina
77. per iment the hydraulic conductivities are usually much higher than the fluxes and the gradients of the hydraulic head VH are therefore small Because of inevitable noise of tensiometer measurements reflecting measurement and calibration error very small gradients cannot be resolved and hydraulic conductivities which are much higher than the flux q cannot be determined with sufficient accuracy To account for this the criterion used by Peters and Durner 2008 may be applied to reject all data that are calculated from gradients smaller than 60 Az where op is the measurement noise of the tensiometer readings Note that in HYPORP FIT 3 0 the default value for the tensiometer uncertainty is set to og 0 15 hPa This means that even small systematic offset shifts in tensiometer readings can lead eas ily to erroneous results and must be strictly avoided If set porperly suctions must differ in the initial phase of an evaporation experiment when almost hydrostatic conditions prevail by exactly 2 5 cm 47 DI HYPROP FIT Software Appendix 2 Automatic estimation of the initial water content After a measurement campaign the initial water content is determined precisely from subtracting all tare weights from the total mass of the measuring assembly 1 e by subtracting the tare weight of sensor unit the mass of the sampling ring and the mass of the dry soil obtained after 24 hours drying at 105 C If this value is not yet available
78. png created gt The group History gives an overview of all recently executed exports 41 HYPROP FIT Software 5 Processing multiple data sets D HYPROP FIT C Users wdurner Dropbox Daten _projektbezogene Daten Laborpraktikum 2014 HYPROP HYPROP 3 JKI bhdix 3 1 12 16476 0 EA File Extra Help www ums muc de po zx i infomation Cod Fitting Q Epot A Soil hydraulic model selection Retention function parameters Conductivity function parameters cigred POI o EME Parameter Value Min Max Unit Parameter Value Min Max Unit kl e Fitting alpha 0 00001 05 Tem Ks 0 01 10000 cm d C F A Optimization Brooks Corey vin 1 01 15 tau 1 10 P E Fredlund Xing 0 C th_r 0 04 cm cm3 v omega 1E 09 0 1 Kosugi C C C C this 0 1 1 cm cm a 1 500 5 0 T T T T pF_dry 6 80 5 00 7 00 van Genuchten m 1 1 n Jim 001 1 GGx bhdix bhdx van Genuchten mnvar C C C C JKO Model FX PDI Model Code 6111 JKI 5 Source not yet published Description PDl varant of the Fredlund Xing model JKI 03 pate K theta Show data as small points i Retention O pF Conductivity K pF data added data function Q Oo vol water content log 10 K in cmd function data added data 6 bd PowerUser HYPROP FIT offers the convenient and powerful option to assemble individual data sets
79. r is fully responsible for data weighting Each data pair will get a single weight No extra calculations will be carried out If the user has prior knowledge about measurement errors this scheme might be used For normally distributed uncorrelated measurement errors with zero mean the single weights can be set to the reciprocal of the variance of the measurement error This is in accordance with the maximum likelihood principle for the method of least squares Omlin and Reichert 1999 Thus in this scheme We and wx are both set to unity and Wg 1 06 and Wi 1 og where og and og are the variances of single the measurement errors In the second scheme the weights are first normalized by a factor for the data type and second by a factor for the data frequency as suggested by Peters and Durner 2008b To account for the different measurement frequency the individual weights wg and Wx are chosen such that the combined data within every log h cm pF increment have the same weight i e the weight for a certain data point is proportional to its distance to the neighboring point on the pF scale To account additionally for the different data types the weights for the data classes are calculated by wg 63 DIZ HYPROP FIT Software 1 Omax Omin and We 1 log 9 Kmax 1084 Krnin where Omax Omin Kmax and Kmin are the maximum and minimum values of the data sets to which the models are fitted The third scheme i
80. rantee both for the parameter t of the Mualem conductivity model Eq 17 the lower allowed boundary of t can be expressed as a function of the pore size distribution parameters of the retention function as described by Peters et al 2011 For the unscaled uni and bimodal van Genuchten retention functions the lower allowed value is given by Tmin 1 2 m and Tmin 1 2 winl1 mil respectively For the 3 2 Integral versus classic fit For Eq 52 the predicted water contents 6 are either calculated in a standard manner as the point water contents at pressure head h classic method or as the mean water content of the column calculated as the integral of the water content distribution over the soil column divided by its height integral method 62 HYPROP FIT Software In the case that the measured water contents and pressure heads are regarded as point measurements in the soil they are treated as simple points of the fitted soil water retention function We call this scheme the classic method In the case of hydrostatic column experiments or when evaporation experiments are conducted and the water contents are measured as mean water contents in the whole soil column with a scale Peters and Durner 2006 2008b introduced the so called integral method This is done since the assumption that the water content is spread out linear over the column is not always fulfilled in coarse pored or structured soil
81. rs A and W Durner 2008a Simplified Evaporation Method for Determining Soil Hydraulic Properties Journal of Hydrology 356 147 162 Peters A and W Durner 2008b A simple model for describing hydraulic conductivity in unsaturated porous media accounting for film and capillary flow Water Resour Res 44 W11417 Peters A S C Iden und W Durner 2015 Revisiting the simplified evaporation method Identification of hydraulic functions considering vapor film and corner flow Journal of Hydrology 527 531 542 Priesack E and W Durner 2006 Closed form expression for the multi modal unsaturated conductivity function Vadose Zone Journal 5 121 124 Schelle H S C Iden A Peters and W Durner 2010 Analysis of the agreement of soil hydraulic properties obtained from multistep outflow and evaporation methods Vadose Zone Journal 9 1080 1091 Schelle H S C Iden and W Durner 2011 Combined transient method for determining soil hydraulic properties in a wide pressure head range Soil Sci Soc Am J 75 1 13 Schindler U 1980 Ein Schnellverfahren zur Messung der Wasserleitfahigkeit im teilgesattigten Boden an Stechzylinderproben Arch Acker u Pflanzenbau u Bodenkd 24 1 7 Schindler U W Durner G von Unold and L Muller 2010a Evaporation Method for Measuring Unsaturated Hydraulic Properties of Soils Extending the Measurement Range Soil Sci Soc Am J 74 1071 1083 Schindler U W Durner
82. s for the tensiometer The uncertainty of the scale 0 05g No Yes readings transposes to uncertainties of the calculated water contents and thus affects the uncertainty estimates for the optimized functions parameters in register fitting Tensiometer Offset Correction Zero Offset Comection Upper Tensiometer hPa 0 00 Lower Tensiometer hFa 0 00 An offset correction is required if the user has an indication from the measurements that the offset calibration of a tensiometer was not correct In general the tensiometers must show under almost hydrostatic conditions which prevail for most soils in the initial measurement phase a difference of 2 5 hPa Deviations from this difference will strongly affect the calculation of the conductivity values at the wet end of the range If a tensiometer time series appears shifted by an offset it is 20 HYPROP FIT Software possible to correct the whole data set for that shift by inserting the offset in the respective box The correction will be directly applied to all recorded tensions and accordingly visualized in the measurement window WARNING Be careful Please correct tensions to assumed hydrostatic conditions only if you are certain that these conditions were actually given at a given time of your measurement campaign Variable Explanation default editable editable value by by public power user user Upper Added offset for the measured values of the upper O hPa No
83. s similar to the second scheme regarding the weights on the pF scale i e Wg and Wx i The weights for the different classes i e Wg and Wx are in this scheme given by the user Schelle et al 2010 for example set w and wx to 1 and 0 001 respectively and found good results Note that the last two weighting schemes are only applicable in the case that no repetitions are measured because then different data pairs may have similar or equal pF values so that several data points will get weights of or close to 0 64 HYPROP FIT Software 4 Post processing 4 1 Diagnostic variables The first measure to evaluate the performance of the fitted model is the objective function values at their estimated minimum min in our case the weighted sum of squared residuals A more descriptive measure giving the mean deviation between model and data is the root mean Square error RMSE 1 J RMSE dui 1LVi 54 where y and are measured and model predicted quantities i e water contents 8 or hydraulic conductivities logio K For a sound representation of the data by the model the values of RMSE should be close to the measurement error The Nash Sutcliffe index Nash and Sutcliffe 1970 and coefficient of determination are calculated for both data types The Nash Sutcliffe index is given by 2 E 1 21lyi i 55 Lilyi il 55 where y is the mean measured quantity The coefficient of determination r is gi
84. sus time the dark blue line the corresponding course of the lower tensiometer The x and y axes are automatically scaled to show all recorded data Visualizing measured datapoints The graph shows by default the temporal evolution of the tensions as lines without individual dots for the data If desired a user can visualize additionally all individually recorded data by moving the cursor on a line right hand clicking the mouse and selecting the respective option from the context sensitive pop up menu De selecting the option leads the individual data points again to disappear Data can also be shown by clicking on the respective on off button in the header line directly above the graph 25 HYPROP FIT Software Visualizing the supporting points used for the smooth interpolation of the data As indicated in Peters and Durer 2008 not all data are used for the calculation of the retention and conductivity data but just a limited number of support data that are taken from an Hermitian spline interpolation of measured data The actual number of these support data can be selected by the user in the submenu Interpolation Options in the register Evaluation see below Support data can be visualized by clicking on the on off button Show Spline Support Points in the header line directly above the graph Search Stop Point Use Air Entry Point Search Air Entry Point Show Spline Interpolation Points Show Data Points Tensio
85. tion with the Burdine conductivity model Burdine 1953 gives Brooks and Corey 1964 2 T 2 K h isle P for h gt a t 44 K for h lt a t As for the Mualem model the slope of the conductivity function vs pF is determined by the pore size distribution parameter of the retention function and additionally by a tortuosity parameter 7 which is regarded as fitting parameter The capacity relation for the Brooks and Corey function is given by 0 0 ah for h gt a C h 45 0 for h lt a 2 2 Fredlund and Xing model The Fredlund and Xing 1994 soil water retention model accounts for zero water content at oven dryness by OCh Osxh i Ch 46 with 59 DIED HYPROP FIT Software P h Infe ah 47 and x h 1 EE 48 where a n m and h are curve shape parameters and e is the Euler number For the Fredlund and Xing retention model there exsist no analytical solution for the Mualem capillary bundle model Therefore equation 17 is solved numerically The capacity function for the Fredlund and Xing retention model is _ ap 4 a C h ma Pak 49 where dx 1 E e 50 dh h hy in 50 and av mn ah in ah e 1 dh n ah eh 51 60 HYPROP FIT Software Tab 1 Summary of all implemented model combinations in SHYPFIT2 0 Basic functions Basic vG van Genuchten constrained Kos Kosugi vGm van Genuchten unconstrained Modality Mod uni
86. tting algorithm minimizes the sum of squares deviations between data points and fitted functions Conductivity data are fitted on a log K scale because otherwise the large conductivity data would completely dominate the fitting result Fitting both data types simultaneously is a multi objective problem and improving the fit for the retention data sometimes can be only accomplished by a worse fit for the conductivity data and vice versa The balance between fitting accuracy for the retention data and the conductivity data is specified by default with a stronger weight for the retention data but the parameters can be modified by power users to any desired weighting scheme see section 4 4 6 4 4 2 Selecting a model and specifying parameter bounds Soil hydraulic model selection Retention function parameters Conductivity function parameters wear PDI ieee Eag _ Parameter Value Min Max Unit Parameter Value Min Max Unit 7 alpha 0 00001 0 5 1 cm Ks 0 01 10000 cm d Brooks Corey Vin 1 01 15 v tau 6 10 Fredlund Xing C C v th_r 0 0 4 cm cm a fy ya m 3 1 E v th_s 0 1 1 cm cm van Genuchten m 1 1 n Sannnnnnnnnnnnnnnnnnnnne van Genuchten mnvar CO C Model VG Model Code 1100 Source van Genuchten 1980 Description traditional constrained van Genuchten Mualem model A hydraulic model is selected by marking the respective tick box By default all model parameters are included in the fitti
87. us time courses of measured data and an automatic detection the validity range of conductivity data in the range near saturation where the hydraulic gradient becomes too small to yield reliable data The simplified evaporation method yields correct results for soils where the water flow is validly described by the Richards equation with time invariant hydraulic properties and where hydraulic properties that can be described with parametric expressions for hydraulic properties including isothermal vapour diffusion Peters et al 2015 For further information about HYPROP FIT Software the scientific background of the method the reader is referred to the scientific publications listed in the references section HYPROP FIT is freely available through the UMS website DI HYPROP FIT Software Z2 Installation of the Software 2 1 Installation using the Microsoft Installer The easiest way to install the software would be to call up the file setup exe directly with the Explorer Double click on HYPROP msi Then follow the instructions on the screen of the installation assistance If you cannot execute the file setup exe you might need a current version of the Windows Installer or and NET Framework 4 0 by Microsoft 2 2 Access rights Public User and Power User HYPROP FIT is run in two user modes which differ in the right to write or change parameters Public Users run the software in a standard manner and apply default valu
88. ven by r2 1 ibi Fil i J 56 To account for the different number of adjustable parameters when comparing different models fitted to the same data SHYPFIT2 0 also calculates the Akaike Information Criterion AIC 2 L k Akaike 1974 where L is the likelihood function and k is the number of fitting parameters In a least squares estimation and with N 0 o7 distributed residuals the information criterion AIC can be expressed in the form as described by Ye et al 2008 AIC nin n 2k 57 where n is the number of measured data and k is the number of adjustable parameters If the number of measurements n is small in comparison to k the original form of AIC should be extended by a correction term that accounts for small values for n k Hurvich and Tsai 1989 leading to the extended information criterion 2k k 1 AlCc nIN min 1 2k n k 1 58 The first term penalizes a poor fit the second term the number of parameters and the third term is the correction term for small values of n k If n k becomes large the last term becomes negligible and the AlCc converges to AIC 65 eas HYPROP FIT Software Since sometimes other information criteria are suggested as well SHYPFIT2 0 calculates also the BIC and KIC for model selection These two are given for least square estimations as follows Ye et al 2008 BIC nin n kIn n 59 and KIC n k In n kln 27 In J wJ

HYPROP-FIT Manual

Contents

Download Pdf Manuals

Related Search

Related Contents