1  Model overview

1.1 Model domain and processes

SWAP simulates transport of water, solutes and heat in the vadose zone in interaction with vegetation development. The model employs the Richards equation including root water extraction to simulate soil moisture movement in variably saturated soils. In addition to ordinary matrix flow, SWAP considers flow through macropores as may occur in clay and peat soils. Solute transport includes the basic processes convection, dispersion, adsorption and decomposition. For more detailed solute transport studies, SWAP can be used in combination with specialized chemical transport models such as PEARL for pesticides and Soil-N or ANIMO for nitrogen. SWAP simulates soil heat flow taking into account actual heat capacities and thermal conductivities. The generic crop growth module WOFOST is incorporated to simulate leaf photosynthesis and plant growth. The soil moisture, heat and solute modules exchange status information each time step to account for their interactions. On a daily basis crop growth is affected by actual conditions of weather, soil moisture and salinity. An extensive test protocol ensures the numerical code quality of SWAP.

Figure 1.1: SWAP model domain and transport processes.

In the vertical direction the model domain reaches from a plane just above the canopy to a plane in the shallow groundwater (Figure 1.1). In this zone the transport processes are predominantly vertical, therefore SWAP is a one-dimensional, vertical directed model. The flow below the groundwater level may include lateral drainage fluxes, provided that these fluxes can be prescribed with analytical drainage formulas. The model is very flexible with respect to input data at the top and bottom of the soil column. At the top in general daily weather conditions will suffice. For Nordic conditions a simple snow storage module has been implemented. In case of more focussed studies (e.g. runoff or diurnal transpiration fluxes) evapotranspiration and rainfall data can be specified in more detail. At the bottom various forms of head and flux based conditions are used.

In the horizontal direction, SWAP’s main focus is the field scale. At this scale most transport processes can be described in a deterministic way, as a field generally can be represented by one microclimate, one vegetation type, one soil type, and one drainage condition. Also many cultivation practices occur at field scale, which means that many management options apply to this scale. Upscaling from field to regional scale for broader management policy studies is possible with geographical information systems.

The smallest time steps in SWAP are in the order of seconds for fast transport processes such as intensive rain showers with runoff or flow in macroporous clay soils. These time steps are automatically increased in periods with more slow flow conditions. Depending on simulation complexity, computation times for 50 year periods range from 30 to 500 seconds on ordinary personal computers.

1.2 SWAP installation

Figure 1.2: Installed folders by standard
SWAP installation.

The SWAP model can be downloaded from Internet site swap.wur.nl. This site contains general information on model features, applications and test reports. Various SWAP versions are available at the Internet site. This manual applies to SWAP 4. Only the most recent SWAP version is supported by the SWAP team.

To allow installation on different operating system, we compiled a versatile set for distribution and packed all files into one zip-file. The zip-file can be unpacked on an arbitrary folder. Once unpacked a number of subfolders are created (Figure 1.2), which contain:

  • Swap executable
  • Swap source code
  • User manual
  • Several case studies
  • Additional input data:
    • Daily weather data of Wageningen meteorological station of the period 1971-2000.
    • Simple crop input data for grass, fodder maize, potato, sugar beet and winter wheat.
    • Detailed crop input data for winter wheat, grain maize, spring barley, rice, sugar beet, potato, field bean, soy bean, winter oilseed rape and sunflower.

SWAP can be automatically launched for the example Hupsel. Also runs can be made with the other 5 case studies described in Chapter 12.

1.3 Model input

The input data of SWAP are divided over 4 different file types:

  • Main input file (*.swp)
  • Meteorological file (*.met, *.rain or *.yyy)
  • Crop growth file (*.crp)
  • Drainage file (*.dra)

Tip 1.1 provides an overview of the information in these input files. The main input file and the meteorological data file are always required. Input files of crop growth and drainage are optional. The extensions of the files are fixed, with the exception of meteorological (and rainfall) data supplied in annual files, for which the extension equals the last three digits of the year (e.g. 2017 yields .017). The names of the input files are free to choose and are specified in the main input file. As listed in Tip 1.1, the main input file contains general information with regard to the simulation, meteorology, crop rotation scheme, irrigation, soil water flow, heat flow and solute transport. For meteorological data, commonly a file with daily data is used. In Chapter 3 also more detailed input of evapotranspiration and rainfall fluxes will be discussed. The detailed crop growth input file is required to simulate crop development and biomass assimilation. As an alternative, the development of crop parameters as leaf area index or rooting depth can be prescribed in the simple crop growth input file. The drainage input file contains two sections. The first, basic drainage section provides input for drainage towards ditches and/or drains. The second, extended drainage section provides input for drainage including simulation of surface water levels.

Tip 1.1: Summary of information in input files. Optional files are denoted with #.

Main input file (*.swp)

  • General section
    • Environment
    • Timing of simulation period
    • Timing of boundary conditions
    • Processes which should be simulated
    • Optional output files
  • Meteorology section
    • Name of file with meteorological data
    • Rainfall intensity
  • Crop section
    • Crop rotation scheme (calendar and files)
    • Crop data input file
    • Calculated irrigation input file
    • Crop emergence and harvest
    • Fixed irrigation parameters (Amount and quality of prescribed irrigation applications)
  • Soil water section
    • Initial moisture condition
    • Ponding
    • Soil evaporation
    • Vertical discretization of soil profile
    • Soil hydraulic functions
    • Hysteresis of soil water retention function
    • Maximum rooting depth
    • Similar media scaling of soil hydraulic functions
    • Preferential flow due to soil volumes with immobile water
    • Preferential flow due to macro pores
    • Snow and frost
    • Numerical solution of Richards’ equation
  • Lateral drainage section
    • Name of file with drainage input data (optional)
    • Name of file with runon input data (optional)
  • Bottom boundary section
    • Name of file with bottom boundary conditions (optional)
    • selection out of 8 options
  • Heat flow section
  • Calculation method
  • Solute transport section
    • Specify whether simulation includes solute transport or not
    • Top boundary and initial condition
    • Diffusion, dispersion, and solute uptake by roots
    • Adsorption
    • Decomposition
    • Transfer between mobile and immobile water volumes (if present)
    • Solute residence in the saturated zone

File with daily meteorological data (*.met, *.rain or *.yyy)

  • Radiation, temperature, vapour pressure, daytime wind speed and/or reference evapotranspiration,
  • Rainfall and/or rainfall intensities

File with detailed crop growth (*.crp) #

  • Crop section
    • Crop height
    • Crop development
    • Initial values
    • Green surface area
    • Assimilation
    • Assimilates conversion into biomass
    • Maintenance respiration
    • Dry matter partitioning
    • Death rates
    • Crop water use
    • Salt stress
    • Interception
    • Root growth and density distribution
  • Calculated Irrigation section
    • General
    • Irrigation time criteria
    • Irrigation depth criteria

File with simple crop growth (*.crp) #

  • Crop section
    • Crop development
    • Light extinction
    • Leaf area index or soil cover fraction
    • Crop factor or crop height
    • Rooting depth
    • Soil water extraction by plant roots
    • Salt stress
    • Interception
    • Root density distribution and root growth
  • Calculated Irrigation section
    • General
    • Irrigation time criteria
    • Irrigation depth criteria

File with drainage data (*.dra) #

  • Basic drainage section
    • Table of drainage flux - groundwater level
    • Drainage formula of Hooghoudt or Ernst
    • Drainage and infiltration resistances
  • Extended drainage section
    • Drainage characteristics
    • Surface water level of primary and/or secondary system
    • Simulation of surface water level
    • Weir characteristics

SWAP uses the TTUTIL library (Van Kraalingen and Rappoldt 2000) for reading input files. Tip 1.2 gives an example of a part of the *.swp input file. General rules for the format of input files are:

  • free format with the structure ‘VARIABLE_NAME’ = ‘value’ or, in case of arrays, in a table with variable names in the first line;
  • order of variables is free;
  • comment in lines is allowed starting with symbols ‘*’ or ‘!’;
  • blank lines are allowed.

The input files list of each parameter their symbolic name, a description and an identification. The identification between square brackets provides information on:

  • the range
  • the unit
  • the data type (I = integer, R = real, Ax = character string of length x) For example: [-5000 .. 100 cm, R] means: value between -5000 and +100 with a unit in cm, given as a Real data type (which means that the value in the input file should contain a dot).
Tip 1.2: Example of input according to TTUTIL in main file *.swp. 
**********************************************************************************
* File name
  METFIL = 'hupsel.met'      ! File name of meteorological data, in case of yearly files without extension .YYY, [A200]
                             ! Extension is equal to last 3 digits of year, e.g. 022 denotes year 2022
                             ! In case of meteorological in one file use extension .met

* Details of meteo station:
  LAT = 52.0                 ! Latitude of measurement site [-90..90 degrees, R, North = +]

* Type of weather data for potential evapotranspiration
  SWETR = 0                  ! 0 = Use basic weather data and apply Penman-Monteith equation
                             ! 1 = Use reference evapotranspiration data in combination with crop factors

* In case of Penman-Monteith (SWETR=0), specify:
  ALT = 10.0                 ! Altitude of measurement site [-400..3000 m, R]
  ALTW = 10.0                ! Height of wind speed measurement above soil surface (10 m is default) [0.1..99 m, R]
  ALTH = 2.0                 ! Height of humidity and temperature measurement (1.5 m is default) [0.1..99 m, R]
  
  ANGSTROMA = 0.25           ! Fraction of extraterrestrial radiation reaching the earth on overcast days (0.25 is default) [0..1 -, R]
  ANGSTROMB = 0.50           ! Additional fraction of extraterrestrial radiation reaching the earth on clear days (0.50 is default) [0..1 -, R]

* Switch for distribution of E and T:
  SWDIVIDE = 1               ! 0 = Based on crop and soil factors
                             ! 1 = Based on direct application of Penman-Monteith

* In case of SWETR=0, specify time interval of evapotranspiration and rainfall weather data
  SWMETDETAIL = 0            ! 0 = Time interval is equal to one day
                             ! 1 = Time interval is less than one day

* In case of daily meteorological weather records (SWMETDETAIL=0):
  FWIND = 1.0                ! Factor for converting 24h averaged windspeed to daytime average windspeed (7-19h; default 1.0) [0.5..5 -, R]

* Switch for use of actual rainfall intensity (only if SWMETDETAIL=0):
  SWRAIN = 1                 ! 0 = Use daily rainfall amounts
                             ! 1 = Use daily rainfall amounts + mean intensity
                             ! 2 = Use daily rainfall amounts + duration
                             ! 3 = Use detailed rainfall records (dt < 1 day), as supplied in separate file

* If SWRAIN=1, then specify mean rainfall intensity [0.d0..1000.d0 mm/d, R] as function of time [0..366 d, R]
  RAINTB =
    1.0  20.0
  360.0  40.0
* End of table

* If SWRAIN=3, then specify file name of file with detailed rainfall data
  RAINFIL = 'hupsel.rain'    ! File name of detailed rainfall data without extension .YYY, [A200]
                             ! Extension is equal to last 3 digits of year, e.g. 003 denotes year 2003
                             ! In case of rainfall in one file use extension .rain

**********************************************************************************

SWAP will read times according to the following format: 2017-02-10_16:30:00.00 denotes February 10, 2017 at 4.30 PM. If in the input only dates and no times are specified, SWAP will assume time 0:00. In case of input data series as function of time (e.g. groundwater levels) or depth (e.g. initial pressure heads), SWAP will apply linear interpolation for times and depths in between. Outside the specified range, the closest value will be adopted. For instance in Tip 1.2, the rainfall intensity (raintb) will gradually increase from 20 to 40 mm d-1 between t = 1.0 and t = 360.0 d. At t > 360.0 d, the rainfall intensity will be 40 mm d-1.

1.4 Model run

The most common way to run SWAP is by executing a batch file, which refers to the SWAP executable and the main input file *.swp. The batch file and the *.swp file need to be present in the same directory. The *.swp file contains the names and locations of other input files. In this way the meteorological, crop and drainage data can be specified on separate folders.

An example of the batch file is given in Tip 1.3. In this case SWAP will use hupsel.swp as main input file. If no name is specified behind the executable call, SWAP assumes swap.swp as main input file. The pause statement keeps the window box with screen messages open; this is convenient when runtime warnings or errors occur.

Tip 1.3: Example of batch file to run SWAP with input file hupsel.swp. 
c:\Program Files\SWAP\swap.exe hupsel.swp
pause

An alternative to run SWAP is by double-clicking file swap.exe. In that case the main input file should be called swap.swp, and should be located in the same folder as file swap.exe.

Three types of messages may occur during a model run:

  • error messages with respect to the input data
  • warnings with the advise to adapt the combination of selected options because the specified combination is not feasible
  • fatal calculation errors which stop the simulation

Output files are created in the same directory as the main input file. The log file, named *.log, contains details on the iteration statistics. If the simulation completes successfully, the file ends with the message: ‘SWAP simulation okay!’. Any warnings generated during the simulation are stored in the file *wrn. If an error occurs, the simulation stops, and the corresponding error message is recorded in the file *.err.

1.5 Model output

Output from SWAP is stored in general ASCII files, which can be read with any editor or word processor. Some output files are always generated, other files are optional. Tip 1.4 provides an overview of the variables that are printed in each output file. All output files have the common header containing the project name, file description, file name, model version and the time of generation. SWAP allows users to define a custom set of output variables and store them in a single file (*_output.csv), which can be easily imported into Excel (see Appendix A for further details). The output interval may range from 0.001 day to 1.0 year. If the output time only consists of a date, the output represents the situation at the end of the particular day. We may distinguish output of state variables and incremental fluxes since the last output time. The output file with final values of state variables (*.end) can be used as input for a subsequent simulation period. This might be useful to derive suitable initial conditions.

Tip 1.4: Summary of information in output files. Optional files are denoted with #.

Short water and solute balance (*.bal) #

  • Final and initial water and solute storage
  • Water balance components
  • Solute balance components

Extended water balance (*.blc) #

  • Final and initial water storage
  • Water balance components of sub systems

Final values of state variables (*.end) #

  • Snow and ponding layer
  • Soil water pressure heads
  • Solute concentrations
  • Soil temperatures
  • Crop status

Output of selected variables (*_output.csv) #

  • Time-series of variables
  • Time-depth-series of variables

Logging of iteration statistics (*.log)

  • Iteration statistics

Warning messages (*.wrn) #

  • Warning messages

Error messages (*.err) #

  • Error messages

In addition to the ASCII files, formatted and unformatted (binary) export files can be generated with data that cover the entire simulation period. These output files can be used as input for advanced solute transport models, such as PEARL (Leistra et al. 2001) for pesticides and ANIMO (Groenendijk et al. 2005) for nutrients. A description of these export files is given in Appendix B.

1.6 Case studies

Chapter 12 contains 4 case studies with SWAP. The input files of these cases come with the installation of SWAP. The case studies are:

  • Hupsel Brook
  • Grass growth
  • Macropore flow
  • Salinity stress

The first case of a field in Hupsel Brook catchment includes a simulation with 3 different crop growth modules. This case is used to illustrate the input files in this SWAP manual.

1.7 Reading guide

In the next chapters we discuss subsequently:

The first part of Chapter 2 describe the physical relations incorporated in SWAP. This part also describes implemented numerical procedures, if required to use SWAP in a proper way. The second part of each chapter describes the model input. If relevant, suggestions for input are included.

The appendices contain information on:

  • User defined output (Appendix A)
  • Formatted and unformatted binary output files (Appendix B)
  • Equations for the implicit linearization of hydraulic conductivities (Appendix C)
  • Equations for the partial derivatives of Fi to pressure heads (Appendix D)
  • Numerical scheme of soil water boundary conditions (Appendix E)
  • Application of the Penman Monteith method (Appendix F)
  • Tables with critical pressure heads for root water extraction (Appendix G)
  • Solution procedure process-based root water uptake models (Appendix H)
  • Derivation and examples of macropore equations (Appendix I)
  • Shrinkage characteristic data (Appendix I)
  • Tables with salt tolerance data (Appendix J)
  • Equations for the numerical solution of heat flow (Appendix L)

1.8 Area of application

The core of SWAP is formed by the general water movement equation in porous media (soil). This is a non-linear, partial differential equation. It has to be solved by numerical mathematics, and the solution is determined by the initial and boundary conditions of the system considered. Technically the implementation of the numerical solution procedure in SWAP has been verified via comparison with analytical solutions (see Chapter 13): from the good correspondence it can be concluded that the core of SWAP is well programmed. Applications of SWAP in practice are fully determined by the user-supplied initial condition and boundary conditions. This is part of the modelling that depends on the user, for which SWAP offers several options and sub-models. The result of the SWAP simulation is highly dependent on the user-supplied information (remember: garbage in = garbage out).

SWAP is primarily meant for simulating one-dimensional situations (vertical water movement), with possibilities of exchange of water with the surroundings via drainage and/or infiltration to/from drainage systems (drain tubes, ditches, canals); SWAP can thus be seen as quasi-three-dimensional. SWAP results should be seen as outcome for the square-meter scale up to the scale of a single field or parcel. For larger areas several SWAP runs need to be performed. Horizontal exchange of water between individual SWAP soil columns is not automatically possible.

For detailed 2D or 3D calculation SWAP cannot be used; for example, the determination of the extent of the wetted bulb underneath a dripper cannot be simulated.

1.9 Major assumptions in SWAP

SWAP is based on the following major assumptions:

  • water movement is dominantly one-dimensional (vertical); quasi-three-dimensional aspect have been included as exchange with drainage systems (drain tubes, ditches, canals; infiltration from these systems into the soil can also be considered);
  • water movement in porous media requires that the porous medium (soil) is rigid, isotropic, isothermal and water is incompressible; swelling and shrinking can be partly considered in case the option macropore flow is used;
  • only single phase flow (water) is considered; air in the soil always remains at atmospheric pressure;
  • soils can be variably saturated;
  • soils can be heterogeneous: the soil then consists of different soil layers each with its own soil physical properties; in principal these are given by the Mualem – van Genuchten equations (other relationships could be provided as tabulated input data); hysteresis can be considered when the Mualem-van Genuchten relationships are used;
  • top and bottom boundary conditions can be either flux-controlled or pressure head controlled; at the bottom boundary several options can be considered; at the top a pressure head condition cannot be supplied by user but only occurs internally when ponding arises;
  • transpiration reduction can be caused due to drought, lack of oxygen (too wet) or too saline conditions; transpiration reduction is used in WOFOST to predict crop growth reduction;
  • ponding at the soil surface occurs when the infiltration capacity is exceed; surface runoff occurs after a threshold ponding height is exceeded and is calculated in analogy to Manning’s overland flow;
  • multiple level drainage systems can be considered to mimic discharge in a sub-catchment; imposed water levels in these drainage systems can be considered;
  • macropore flow can be considered as a special case (e.g. for structured soils such as clay and peat); shrinkage and swelling of these macropores can be included;
  • the crop is explicitly simulated in SWAP via incorporation of the WOFOST model for major arable crops (maize, potato, sugar beet, winter wheat, (summer) barley); for grassland a detailed crop growth module was derived according to the same philosophy as WOFOST; other crops can be simulated via a simple crop growth module;
  • solute transport is described by the dispersion-diffusion – convection equation including simple transformation equations, adsorption and root uptake; for more detailed solute transport coupling of SWAP with ANIMO (Groenendijk et al. (2005); Renaud et al. (2005) is suggested; a simple single-layer N-module is present (Groenendijk et al. (2016));
  • soil temperature can be simulated as a diffusion process; soil temperatures can influence crop growth, solute transformations etc.;
  • simple snow and frost options are available (requires soil temperature);
  • irrigation can be considered either as fixed times and quantities, or chosen from five trigger options and three irrigation depth (amount) criteria.

1.10 Summary of fitness-for-use

SWAP can be downloaded from the website (swap.wur.nl) and comes as a zip-file. This file can be unzipped to any folder. No formal installation procedure is required (no administration right is needed). Standard, SWAP is available as an executable for Windows platforms. The source code is available, so that users are able to produce their own executables e.g. for Linux platforms; from own tests we know that SWAP can be compiled and run under Linux without changes in the source code.

SWAP is not a simple, graphically oriented plug-and-play application. The user must supply information in ASCII input files and then run SWAP by hand indicating the name of the main input file (*.swp); see general description Section 1.4. After completion of the model run, the results are available in several output ASCII files. The user can use these in their own scripts for further analyses.

Basically, based on meteorological information, crop parameters, soil properties, initial and boundary conditions SWAP predicts as a function of time the different water balance components in this system. It is the user that supplies all the necessary information, which will be used by the in SWAP implemented theories to predict the fate of water. Output is available as a function of time which can be further analysed by the user.

SWAP is also used in Wageningen University courses. For this a specific graphical user interface was developed so that students can easily change major input information and visualize major output information. Since not all functionalities of SWAP have been incorporated in this GUI, this interface is not distributed together with SWAP download.

1.11 Brief critical analysis of possible shortcomings

The SWAP model is basically a calculator that solves a partial differential equation based upon input information (initial and boundary conditions) as supplied by the user. Technically, the core of the model has been verified against analytical solutions (see Chapter 13). This means that the quality of the output for a user-defined situation is then mainly determined by the quality of the input as supplied by the user (garbage in = garbage out). This is something that cannot be controlled or checked by the model code. The only check performed by the model is done by checking the input values against pre-defined ranges in the source code.

SWAP is basically a point-model in space and only considers depth along which variable soil conditions can be considered. Real soils are heterogeneous, soil water movement may sometimes be 2- or 3-dimensional (e.g., underneath drip irrigation), and soils and their properties may change in time (e.g., swelling-shrinking; compaction). These aspects cannot be considered in a single SWAP run.

See further the description provided on major assumption (Section 1.9) and fitness-for-use (Section 1.10).

1.12 Brief global uncertainty analysis

As mentioned above, SWAP is basically a simulation model for one-dimensional problems (a point model in terms of surface area). By adding the possibilities of side boundary conditions mimicking possible exchange with the surrounding via different levels of drainage/infiltration boundaries, the SWAP application can thus be extended to a field or parcel. However, the outcome still refers to a vertical distribution of state variables. As such, it cannot predict horizontal variation within the field. Comparison with measured field data (e.g., groundwater level, water content, crop yield) can show a mismatch since, for example, the groundwater level gauge may have been installed at a position that is not comparable with the (average) information that was used as input in the model. This means that uncertainty in model outcome is likely present.

Below a brief and global description regarding uncertainty is provided. Such an analysis should focus on the following aspects: context, model, inputs, parameters, output (Walker et al. 2003; Janssen et al. 2003).

Context

SWAP is used on a diversity of projects and research questions with focus on the water balance in the unsaturated-saturated top part (down to one to a few meters) of soils, often in relation to agricultural crop production. Often, these involve studies where geographic differences need to be studied. For that relevant information needed as input should be available at the scale of interest. In order to simulate a region this means that several stand-alone SWAP simulations are needed which are representative for a small area of the total region of interest. The geographical variability in the outcome is then purely driven by the input information. SWAP is a pure technical model and does not include economics, social and political aspects.

Model

Technically, the numerical solution for water movement, solute transport and soil temperature has been verified against analytical solutions, and validation against field data has been performed (see Chapter 13). This, however, does not guarantee certainty in overall SWAP model outcome.

Over the decades the structure of the model has evolved and likely there is a good logistics in exchange of information between the different modules and the sequence in which these modules are used. Several modules are present in SWAP but not all are as sophisticated as others.

In scientific community different descriptions exist for different processes, indicating that there is no universal description available. In SWAP some processes have different alternatives that can be used, either i) depending on data availability or ii) on desired details for which a subprocess needs to be considered. As an example for i): for crop growth it is advised to use the WOFOST module for those types of crops for which default WOFOST data files are available; when other crops need to be considered SWAP offers a simplified crop growth module. As an example for ii): root water uptake is often modelled via the so-called Feddes reduction model in SWAP, but alternative root water uptake models are available that take into account the root-soil interface. One such detailed root water uptake model is already implemented in SWAP and another one will be implemented soon.

There is an option to include snow and frost in SWAP. This is done in a simple way, as these processes hardly play a role for (current) Dutch conditions. A detailed macropore option is present in SWAP. It is not used often, since it requires many input parameters for which data are lacking. Soil temperature can be simulated, and is required in case solute transport or the N-module is considered. Soil temperature is dependent on soil volumetric water content, but heat transport via moving water is not included.

Inputs

The outcome of a SWAP simulation is highly driven by the quality of the input information. In the Netherlands, modellers (users) likely will make use of information obtained from the Dutch soil map, the Staring series (Heinen et al. 2020) or the derived soil physical units map BOFEK2020, which is derived from the soil map and the Staring series (Heinen et al. 2021). Some crop parameters can be obtained from the standard WOFOST database, but for other crops this information must be obtained via expert judgement.

Parameters

In SWAP most of the information needed to perform the simulation is based on information provided by the user (via text input files). In some cases universal or literature constants (parameters) have been hard coded (mostly as parameters in the top of the routine/function/module). For official releases the source code is available and then these parameters might be changed if needed. In the future we might consider to have all hard coded parameters listed in a special parameter input file. This, however, only provides more flexibility, and will not automatically lead to more certain outcome.

Model outcomes

Many outcome variables are available, and users are able to indicate what information they want as output. The output itself comes as is, i.e., no qualitative or quantitative information on the reliability or uncertainty of the outcome is given.

All-in-all, SWAP-WOFOST fulfills the need of having a simulation tool that can consider the water balance in the combined soil-crop domain. It makes use of universally accepted process descriptions, and it is widely used both nationally and internationally, indicating that it is an accepted tool for this type of research problems. Despite this acceptance as well as some shortcomings in the model, we are aware that the modelling outcome comes with uncertainties. Such shortcomings are less likely the result of errors in the numerical implementation (solution) of the governing differential equations, and are more due to uncertainties in the available input information. Therefore, it is strongly advised that for each new study (application) the user should perform some kind of sensitivity analysis to determine where possible problems or uncertainties in outcome may arise.