HYDRUS: 11-20

FAQ 11-20

11. Why may water fluxes in the v.out and Boundary.out files differ?

There may be some (hopefully relatively small) differences between velocities given in the v.out output file, which are drawn in the graphical output, and those printed to the Boundary.out output file.

• The velocity vectors from the v.out file are calculated based on eqs. 5.28 or 5.29 of the Technical Manual. These fluxes are obtained by calculating the arithmetic average of the gradient in a particular node (from all surrounding elements) and by multiplying this gradient by the nodal conductivity. Thus, these velocities are secondary to the numerical solution of water flow and do not affect it (they are used by the solute transport module and thus do affect the solute transport). Because of the numerical technique applied (linear finite element), some errors are always associated with these calculations.

• Velocities printed to the Boundary.out output file are calculated directly from the Q (nodal flux) term in eq. 5.17 by dividing this value by the boundary length associated with a particular node. Since this term (Q) is part of the primary numerical solution, velocities calculated in this way are more precise than those based on 5.28. Boundary fluxes calculated this way are used in HYDRUS in the mass balance calculations.

Conclusions: For further mass balance analysis, one should use values (both fluxes and velocities) from the Boundary.out output file since they are more precise.

12. What is the difference in spatial discretization for groundwater and vadose zone problems?

One cannot have finite elements for variably-saturated flow problems as large as for saturated problems (e.g., MODFLOW, as its users are used to). While one can have large elements in the saturated zone (where the numerical solution is often based on the linear Bousinesque equation), one cannot have such elements in the unsaturated zone (where the numerical solution is usually based on the nonlinear Richards equation), especially close to the upper boundary where one can often encounter sharp infiltration fronts (such as for infiltration into dry soils). There, the discretization must go down to a few cm (or less); otherwise, the numerical solution cannot capture the shape of the infiltration front and often does not converge. The Darcy-Buckingham law is a local equation, and it simply is not valid at larger scales (over large elements). One cannot, for example, have one point at the surface and another 3 m deep and expect that the code can somehow formulate Darcy-Buckingham's law over this distance. First, one should use our one-dimensional model (HYDRUS-1D) to quickly determine what discretization one can use for specific soil hydraulic properties and boundary conditions and then use this discretization for two- or three-dimensional transport domains.

If you are unsure on how to discretize time and space for HYDRUS calculations, read FAQ 3 or download brief Notes on Spatial and Temporal Discretization.

13. What are the most commonly used bottom boundary conditions?

Free Drainage: This is a boundary with a unit pressure head gradient that should be used at the bottom of the soil profile where the groundwater level is very deep and thus does not affect flow in the transport domain. This boundary condition only considers gravitational flow at the bottom of the soil profile.

Deep Drainage: This BC relates flow at the bottom of the soil profile to the position of the groundwater (that must be above the bottom of the soil profile, i.e., within the transport domain). A special function relates this deep recharge with the position of the groundwater level. This function should predict that deep recharge is larger when groundwater is high and smaller when it is deep. You need to have information about this relationship before you can use it.

Seepage Face: This BC should be used on the sides of dikes (where you expect that water may leave the transport domain), on the bottom of lysimeters, or for tile drains. This BC states that there is no flow as long as the boundary is unsaturated, and flow starts only with saturation. The boundary head is set to zero at saturation, and the code calculates the flux.

Variable Pressure Head: This BC should be used when the position of the fluctuating groundwater level is known.

No Flow: This BC should be used when an impermeable soil layer is at the bottom of the transport domain.

14. How to edit manually HYDRUS (version 5) or HYDRUS (2D/3D) input files?

1. First, you need to create HYDRUS input data files (from information stored in the HYDRUS GUI project) using the command "File->Import and Export->Export Data for HYDRUS Solver in Text Format". This is needed since several input files used by HYDRUS are in binary format and cannot thus be manually edited by the user.
2. Second, manually edit the HYDRUS input file you wish to edit (e.g., Selector.in - most flow and transport parameters, Domain.dat - initial conditions and spatial distribution of variables, etc) while keeping the HYDRUS GUI open.
3. Third, import the modified information back to the HYDRUS GUI using the command "File->Import and Export->Import Data from *.in Files".
4. Fourth, check the imported information in the HYDRUS GUI.

15. How to copy x-y line graphs to other applications?

There are two approaches to how this can be accomplished. In each case, adjust first all settings (e.g., caption, text on axis, fonts, line thicknesses, and colors) in the graph as required. Then either:

1. click the right mouse button when having a cursor above the graph, and select the Copy command from a pop-up menu that appears. The graph's content will be copied to the clipboard for later paste (Paste of Ctrl+v) in various other Windows applications (e.g., MS Word, PowerPoint, or Excel).
2. Alternatively, hit the keyboard keys Alt and PrtScr (Print Screen) to copy the entire dialog window to the clipboard. You can then paste it into various Windows applications again.

16. How to copy/send HYDRUS projects to another computer/user?

1. Open the Project Manager in the HYDRUS GUI (in either HYDRUS, HYDRUS-1D, or HYDRUS (2D/3D)).
2. Click on the Project Group Tab (in the upper left corner of the Project Manager window). On this tab, you can see various Project Groups you have and the paths where these Project Groups and individual Projects are located.
3. Using Explorer (or any other file-managing software), go to the location (path) where your projects are located.
4. In HYDRUS-1D, you must send both the Project_Name.h1d file and the Project_Name folder (which contains the project's input and output files).
5. In HYDRUS (Version 5) or HYDRUS (2D/3D), you need to send the Project_Name.hyd5 (h3d3) file when the Working Directory of a particular project is temporary. When the Working Directory of a particular project is permanent, you must also send the Project_Name folder (which contains the project's output files).
6. It is preferable to compress (using the ZIP or RAR format) required files into a single file before sending it to another user.

17. What are the values for various nitrate root uptake parameters?

Dr. Adrian Zammit from Simmonds & Bristow has provided us with a table of cMin, Imax, and Km parameters for many different crops. This table is based on an extended literature review. These parameters are needed when modeling active nitrate uptake.

18. What does it mean when my output shows NaN?

NaN simply means Not a Number. For explanation, see, for example, Wikipedia.

It is likely a result of a wrong definition of the problem, numerical nonconvergence, division by zero, or some similar problems. Most likely, you do not have sufficient spatial or temporal discretization. See my notes on discretization.

This may happen when the transport domain is quite large and not sufficiently discretized. I would recommend starting with a smaller domain and finer discretization, and only once that works, then I would increase the domain.

19. How to display the HYDRUS animation in the PowerPoint presentation?

The animation files (*.avi) created by the HYDRUS command Create Video File can be displayed using standard video software, such as the Windows Media Player. Animation files can also be inserted directly into PowerPoint presentations using the menu command "Insert->Movie->Movie from File …" and selecting whether animation starts Automatically or When Clicked. Animation can then be stopped and restarted using additional mouse clicks.

20. Nonequilibrium solute transport

HYDRUS codes can simulate either chemical nonequilibrium or physical nonequilibrium solute transport. It can handle both physical and chemical nonequilibrium transports simultaneously only in the Dual-Porosity Model with Two-Site Sorption Model in HYDRUS-1D or in the Dual-Permeability Models.

Chemical nonequilibrium transport is implemented using a One-Site Sorption Model (Frac=0), a Two-Site Sorption Model (0 < Frac < 1), or the Two Kinetic Site Sorption Model. In the Two-Site Sorption Model, one fraction of sorption sites have kinetic and another instantaneous sorption. In the Two Kinetic Sites Sorption Model, both sorption sites have kinetic sorption (attachment/detachment).

The physical nonequilibrium model uses the mobile-immobile water concept. HYDRUS decides which of the two nonequilibrium models to use based on the input solute transport parameters. When the immobile water content (ThImob) is specified with a nonzero value, then HYDRUS uses a physical nonequilibrium model, and the Frac parameter is interpreted as a dimensional fraction (between 0 and 1) of the sorption sites in contact with mobile water. When the immobile water content (ThImob) is specified with a zero value and the "Frac" parameter is not equal to 1, then HYDRUS uses a chemical nonequilibrium model. Then, the Frac parameter is interpreted as a dimensional fraction (between 0 and 1) of the sorption sites classified as type-1, i.e., sites with instantaneous sorption. The rest of the sorption sites (1-Frac) are classified as type-2, i.e., sites with kinetic sorption. For both models, the Alpha parameter represents the first-order mass transfer coefficient; either for nonequilibrium adsorption when chemical nonequilibrium is considered or for mass exchange between mobile and immobile regions when physical nonequilibrium is considered.