The specific processes that SUTRASET considers on top of USGS SUTRA are:

1. evaporation taking water away with salt left in the soils
2. the reduction of evaporation due to water desaturation on the surface, by considering surface resistance
3. the reduction of evaporation due to increase to solute concentration on the surface, according to Kelvin Equation
4. the reduction of evaporation due to salt precipitation on the surface as crust, by considering salt resistance
5. the removal of residual liquid water due to evaporation (by extending SWCC and Kr function below residual liquid water saturation level)
6. a dynamic soil surface boundary conditions with three phases (1. Hydrostatic pressure boundary for water submerged surfaces, 2. Seepage for saturated soil surface above water level, 3 evaporation for unsaturated soil surface above water level). These three phases are determined using non-linear iterations to avoid phase lag (water spraying truck effect!).

Vapor flow in porous media is not considered in SUTRASET

the result of a 1-D calibration case considering evaporation from a soil column with a saline water table at the bottom:

The case is available at examples/Fujimaki/MassarLoamy and /examples/Fujimaki/Toyoura

Salt transport in a coastal wetland considering tide, seepage and evaporation

The mathematical explanation is given in the paper:

Shen, C., Zhang, C., Xin, P., Kong, J., & Li, L. (2018). Salt Dynamics in Coastal Marshes: Formation of Hypersaline Zones. Water Resources Research, 1–18. https://doi.org/10.1029/2017WR022021

the case is available at examples/evaporation_salt_marsh/sandyloam_evt4

Development of freshwater lens in an alluvial floodplain considering fresh surface water (on the left), saline groundwater (on the right) and evaporation (on the top)

Reference for example 3:

Packaged input file is downloadable from here.:

America, Ilja, Zhang, Chenming, Werner, Adrian D. and Zee, Sjoerd E. A. T. M. (2020) Evaporation and salt accumulation effects on riparian freshwater lenses. Water Resources Research, . doi:10.1029/2019wr026380

Issue the following command

git clone https://github.com/chenmingzhang/sutraset
cd sutraset/src
make 
make all

You will have sutraset_gf compiled in the src folder

SUTRASET has been pre-compiled as sutraset_gf_windows.exe in the directory bin. The associated dlls are also included in bin folder. This exe file has a alias (or shortcut) in folder bin named sutraset.bat that can be copied to any of your simulation folders. (We do not suggest copying the exe (sutraset_gf_windows.exe) and associated dlls file into the simulation directories for running SUTRASET.)

To allow sutraset.bat to be copied and executed in any of your simulation folders, you will need to add the absolute address of directory bin into the system environment variables.

For example:

If sutraset is located in d:\sutraset, the absolute address of directory bin is d:\sutraset\bin. You will need to add d:\sutraset\bin in to your system environment variable.

The way to add path into system environment variables can be found from here:

https://superuser.com/questions/949560/how-do-i-set-system-environment-variables-in-windows-10

After this, you will be able to copy sutraset.bat into any of your simulation folders and run the simulation by double clicking the file.

After downloading the package in zip file, create a file in folder /src called gversion.f90 with following content

SUBROUTINE GVERSION (K3)
  WRITE (K3,*) "this is not git controlled "
      RETURN
END SUBROUTINE

Then issue the command below in the folder src (with mingw)

gfortran *.f90 *.f -O3 -o sutraset_gf_windows.exe

SUTRASET treats evaporation as a sink term. The node numbers needs to be listed DATASET 17. As the change of evaporation rate due to surface saturation, concentration and solid salt depth is implemented in subroutine BCTIME, these node number will be specified as negative values. Since the original SUTRA does not automatic calculate the effective surface area associated with the flux in source and sink term (dataset 17), User needs to input a volumetric inflow/outflow (m3/s) by multiplying the in/out flux (m/s) with the injecting area (m2), Similarly, SUTRASET requires user to input the node geometry where evaporation is implemented. This is done by to specifying the surface area and the depth of the cell where evaporation is applied, after the negative node number. Please referred to example examples/evaproation_salt_marsh/sandyloam_evt4/evapotranspiration_sandyloam.inp The soil water retention parameters, air conditions, and resistance parameters are listed in input file ET.inp. See the same example mentioned above.

1. If negative solute concentration is present, particularly below the cell where evaporation is implemented, try to increase the diffusivity, or refine the mesh. As evaporation may lift the salinity up to solubility limit (~265 PPT for NaCl), generating large concentration gradient near the surface. This could be relieved by quickly diffuse the hypersaline water through diffusion, or add more cells to describe the details

2. non-linear iteration must be enabled, when seepage is considered.

3. if tide, seepage and evaporation is considered on the top surface, (Refer to example /examples/evaporation_salt_marsh ) and simulation is crashed with P equal NaN, trying to adjust GNUP to minimise the recharge in to surface unsaturated zone. This phenomenon occurs in particular when soil has high permeability. As evaporation may de-saturate the soil surface with a very low saturation (0.01), the corresponding pressure could be -1e7 Pa according to soil water retention curve. Once tides comes with a PBC of 0 Pa, large pressure deficit (in this case 1e7 Pa) could cause significant recharge (e.g., flow_in=GNUP*(PBC-P)=0.01*1e7=1e5m3/kg) and so blow the surface cell.