SWIMv1/SWIMv2 - soil water infiltration and movement model - simulate soil water balances
SWIMv1/SWIMv2 Categories: flow models - unsaturated zone, soil, solute transport models - unsaturated zone
SWIMv1 - Soil Water Infiltration and Movement (Version 1)
SWIMv1 is a software package for simulating water infiltration and movement in soils. As in the real world, SWIMv1 allows addition of water to the system as precipitation and removal by runoff, drainage, evaporation from the soil surface and transpiration by vegetation.
SWIMv1 helps researchers and consultants understand the soil water balance so they can assess possible effects of such practices as tree clearing, strip mining and irrigation management.
SWIMv1 is valuable for scientists and consultants involved in land planning and land management. For example, if a development is being considered which involves tree clearing, SWIMv1 can be used to indicate salinity or surface runoff problems that could result from a change in the soil water balance associated with the removal of the trees.
SWIMv1 is an easy-to-use menu-driven system ideal for teaching students about the soil water balance. With the aid of a computer, students can simulate real-world situations and gain a feel for how the systems behave. They can investigate particular systems such as how water movement through a clay soil differs from water movement through a sandy soil, given the same rainfall conditions. This helps to provide a quick and accurate understanding of various water systems and the effects that different management practices have on the soil water balance.
SWIMv1 consists of a menu-driven suite of three programs that allow the user to simulate soil water balances using numerical solutions of the basic soil water flow equations.
The SWIMv1 simulation program obeys the basic physical law of conservation of mass while making the following assumptions:
While a working knowledge of soil physics and computing offer an advantage in using SWIMv1, efforts have been made to provide a user-friendly package with extensive help facilities. The three main modules comprising SWIMv1 are:
1. SWIMFILE - input data preparation program
2. SWIMEV - the simulation program
3. SWIMPLOT - output data presentation program
SWIMFILE is a menu-driven program which allows the user to edit an input data file interactively. Both the values and units of parameters can be altered, with error flagging occurring for incorrect editing.
Parameters that can be altered include those describing:
Simulation - starting time, finishing time, print interval (simulation results will be output at this interval) and water increment (simulation integration time steps are chosen so that the greatest water flux in the system, excluding transpiration, will move approximately this amount of water in one time step; a smaller increment gives better accuracy).
Vegetation characteristics - the model allows up to four vegetation types to grow together. Each type has certain characteristics that determine its water extraction pattern. A fraction of the total potential evapotranspiration (PET) is assigned to each type, with any left over being used for soil evaporation. The fraction of the potential that a type intercepts and its root length density increase sigmoidally with time. The distribution of roots with depth is assumed to be exponential, with a fixed depth constant.
Soil surface conductance - initial soil surface conductance, minimum soil surface conductance, precipitation constant and effectiveness parameter. Although SWIMv1 can cope with thin impeding layers, there is special allowance for those at the surface. A surface seal or crust can be represented as a surface conductance, and its value may decrease exponentially with cumulative rainfall energy. Alteration of these parameters will alter the surface conductance characteristics of the soil.
Surface storage - initial soil surface storage, minimum soil surface storage, precipitation constant. Runoff occurs when the surface-water depth is greater than the surface storage. To allow for a reduction of surface roughness due to rainfall, the storage decreases exponentially with precipitation energy from the given initial value towards the given minimum in exactly the same way as the surface conductance. Variable initial surface-water depth allows the simulation to begin with the surface ponding.
Runoff - runoff rate factor and runoff rate power. For a water depth that is dh above the storage, the runoff rate is equal to the given runoff rate factor times dh^P, where P is the given runoff rate power. This crudely simulates overland flow.
Soil properties - hydraulic properties together with soil depths and initial matric potential values. The hydraulic properties available include thetas (water content at field concentration), psie (air entry potential), b (minus the slope of a straight line approximating the water retention curve on a log - log plot), and ks (hydraulic conductivity at field saturation). Soil depths are those used for the simulation. Increasing the number of soil depths increases accuracy but at the expense of simulation execution speed.
Precipitation and potential evapotranspiration - cumulative precipitation and cumulative PET are given functions of time. SWIMEV uses linear interpolation to find values between those given in both cases.
SWIMFILE offers extensive help screens providing background information, examples, suggested values and assistance with general editing procedures. A data file is created on completion of this program which is then used as an input file to the simulation program, SWIMEV.
SWIMEV solves Richards' equation numerically using recently-developed techniques that ensure efficient computation (Ross, P.J. 1990. Efficient numerical methods for infiltration using Richards' equation. Water Resources Research 26(2), 279-290). The methods ensure that mass is conserved, even when fast, approximate solutions are obtained. Although Richards' equation does not accurately describe every flow situation, it is the accepted basis of soil water flow. Within its limitations, a user can simulate infiltration, redistribution, deep drainage, simultaneous evapotranspiration by up to four types of vegetation, transient surface-water storage and runoff. Soils may be vertically inhomogeneous, but must be horizontally uniform. Macroporosity can be included if water in the macropores can move quickly to other pores.
SWIMEV accepts data input from an ASCII file created by the data preparation program, SWIMFILE, on a standard text editor.
Output from SWIMEV is written to a binary file, to the screen, and optionally to an ASCII file. The binary output file can be viewed graphically using SWIMPLOT. The FORTRAN source code for SWIMEV is printed as an appendix to the manual.
SWIMEV Data Input
Optional parameters include:
Help screens can be used for guidance in selecting values for entries where there is uncertainty.
SWIMEV Data Output
SWIMPLOT pictorially displays the data from SWIMEV for each print step. VGA color displays are supported.
SWIMPLOT displays a plot of volumetric water content with depth for available and unavailable water, evaporation rates, and water balance components relative to the overall water budget.
The user has control over the display speed and the facilities to restart the display at any screen number, alter increments between screens, pause the screen and print an individual screen. Printer support is provided for Epson-FX and Hewlett Packard Laserjet II and compatible printers.
SWIMv2 (Soil Water Infiltration and Movement model version 2) is a mechanistically-based model designed to address soil water and solute balance issues associated with both production and the environmental consequences of production. SWIMv2 employs fast, numerically-efficient techniques for solving Richards' equation for water flow and the convection-dispersion equation for solute transport and is suitable for personal computer applications. The model deals with a one-dimensional vertical soil profile which may be vertically inhomogeneous, but is assumed to be horizontally uniform. It can be used to simulate runoff, infiltration, redistribution, solute transport and redistribution of solutes, plant uptake and transpiration, evaporation, deep drainage and leaching. The physical system and the associated flows addressed by the model are shown schematically in Fig.1. Thermally-induced vapor flow and the temperature dependence of liquid water movement are ignored. Soil water and solute transport properties, initial conditions, and time dependent boundary conditions (e.g., precipitation, evaporative demand, solute input) need to be supplied by the user in order to run the model. SWIMv2 uses a self-adjusting time step to meet specified error criteria, and the user need only specify the maximum time step allowed. There is no real limitation to the space and time steps that can be used but they should be chosen to reflect the particular processes being studied and the questions being addressed. It is common to use one hour as the maximum time step when addressing diurnal issues, and one day for most other cropping issues that are likely to be addressed with SWIMv2. The option of both going higher (weeks, months, years) and lower (minutes, seconds) is available to the user. The SWIMv2 model conserves mass, no matter how large the space or time steps that are used in the simulation. Additional details of the SWIMv2 model and its use are included in the User Manual (Verburg et al., 1996).
Fig.1. Schematic showing the soil-plant-atmosphere system modeled by SWIMv2.
SWIMv2 is based on a fundamental description of the conservation of mass (water and solutes) within soils and allows the user to address a wide range of issues associated with the soil water and solute balance. The major features of the model which provide this capability include the ability to deal with:
The key soil hydraulic properties required by SWIMv2 in order to solve Richards' equation are the storage (soil water retention curve) and transmission (soil hydraulic conductivity function) properties. SWIMv2 offers the user considerable freedom of choice in terms of the type of function that can be used to describe these properties. For the hydraulic conductivity, these include: the Mualem model, the Brooks-Corey model, the van Genuchten model, the sum of simple functions, and tabulated data.
For the water retention function, these include: the Brooks-Corey function, smoothed Brooks-Corey, modified Brooks-Corey, modified smoothed Brooks Corey, van Genuchten, exponential, sums of simple functions, and tabulated water retention data described in terms of simple polynomials. Hysteresis in the soil water retention function can also be included if required.
A program called HYPROPS is used to generate the required hydraulic property tables used by SWIMv2.
In order to apply the equation for reactive solutes, one needs to specify both the adsorption isotherm (SWIMv2 allows the Freundlich isotherm) and the dispersion coefficient. For nonreactive solutes, one only needs the dispersion coefficient. Accurate determination of these solute transport properties is not trivial.
SWIMv2 can accommodate up to four vegetation types in any one simulation. This enables one to address issues associated with single crops, intercropping or mixed species (trees and grasses). The time trends in potential evaporation for each vegetation type and root density distributions are the main information required, and there is some flexibility in terms of how they are supplied (as simple functional forms or as discrete data points in time).
Potential evapotranspiration is used to determine the evaporative demand placed on the system and needs to be supplied by the user. A fraction of the total potential evapotranspiration is assigned to each vegetation type with any remaining acting as the demand on the soil. The potential demand is exerted on the root system and simple electrical circuit analogies used to determine root water uptake for each soil layer. These calculations are based on steady-state radial flow to roots and take into account the soil and root resistances. If water supply is unable to satisfy the demand, an iterative procedure is followed to determine the actual water that can be taken up as transpiration.
A similar approach is used for evaporation from the soil surface. If the surface is wet and supply can meet demand, evaporation takes place at the potential rate. Once the soil starts to dry and the atmospheric demand cannot be met, actual evaporation is governed by the soil's ability to supply water to the surface.
It is important to note that while SWIMv2 can include up to four types of vegetation, it is not a crop growth model. There is no feedback between the plant and soil processes, and care will therefore be needed when applying SWIMv2 in a stand-alone form. This limitation can be overcome by using SWIMv2 in a cropping system framework such as APSIM. This capability is now available.
Various management options are available to the user including: cultivation on a regular basis over consecutive years (cycling) or at specified times where various parameters such as surface conductance and surface roughness are reset, and adding fertilizer or other solutes at specified times.
There is also an ability to make use of 'discontinuities' to examine effects of major impacts such as might occur following harvesting on very wet soils, or deep ripping. To achieve this, the simulation can be stopped at a specified time and then restarted with new soil property values to reflect the 'new' soil conditions.
Input required by SWIMv2 is used to specify the site and conditions for the simulation and include the initial conditions, boundary conditions, soil hydraulic properties, soil solute properties, vegetation characteristics, management practices, cumulative precipitation (rainfall + irrigation), and cumulative potential evapotranspiration.
The two key boundaries within the system of interest are the top or upper boundary taken as the soil surface, and the bottom or lower boundary taken as some specified depth in the soil below which water and nutrients are considered to have been leached and therefore inaccessible by plant roots. Conditions at these boundaries need to be specified so that the numerical solution procedure works, and the user has a wide range of choices for this. At the top boundary, the user can specify infinite surface conductance, a constant potential (such as occurs when using a disk infiltrometer), specified surface conductance function, surface ponding, no ponding (i.e., any water that cannot infiltrate runs off immediately), or a specified runoff function. At the bottom boundary, one has the choice of gravity drainage, specified gradient, specified water potential (e.g., zero for a water table surface), zero flux, or free drainage (such as occurs during dripping from large cores exposed to atmospheric pressure).
The key output provided by SWIMv2 is the soil water potential, soil water content, solute concentration, and associated fluxes as a function of depth and time. In addition to this output, both instantaneous and cumulative values of transpiration, evaporation, infiltration, storage, drainage, leaching, and plant uptake are provided.
The output of SWIMv2 consists of the output to the screen and a binary output file. The screen output can be redirected to a file if need be. The binary file data can be viewed using the SWIMPLOT program or converted to ASCII tables with the SWIMREAD program and the data then used in any one of a range of analysis and/or plotting packages.
An example output obtained using SWIMPLOT is given in Fig.2.
Fig.2. Example SWIMv2 output displayed. Shows cumulative soil water balance components and soil water content and solute concentration as a function of depth.
PC with 640K RAM
Scientific Software Group
P.O Box 708188 Sandy, Utah 84070
Phone (801) 208-3011 or Toll Free (U.S.) 1-866-620-9214 Fax (801) 302-1160 E-mail email@example.com
Copyright 1998 Scientific Software Group All rights reserved