ARS | Publication request: Analysis of Field-Scale Salinization: a Comparison of Different Modeling Approaches

Submitted to: Journal of Hydrology and Hydromechanics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: February 8, 2005
Publication Date: August 1, 2005
Reprint URL: http://www.ars.usda.gov/SP2UserFiles/Place/53102000/pdf_pubs/P1958.pdf
Citation: Vaughan, P.J., Shouse, P.J., Suarez, D.L., Van Genuchten, M.T., Ayars, J.E. 2005. Analysis of field-scale salinization: a comparison of different modeling approaches. Journal of Hydrology and Hydromechanics. 53(2):73-90.

Interpretive Summary: The transport of salts in soil is of practical interest because the buildup of salts in the root zones of crops can adversely impact productivity. This paper is concerned with computer modeling of the vertical movement of salts, specifically the chloride ions in salts, in an agricultural field in the San Joaquin Valley of California. Calculations of salt movement were made for a two-year period at 45 sites within the field. The results of the calculations were compared with measurements of the amount of chloride in soil samples taken at the end of the period. At 30% of the sites the measurements showed that chloride increased with depth to a maximum at around 1 meter depth and then decreased at greater depths. The model was unable to accurately predict the chloride maxima at these sites even when special improvements were made to handle fast and slow-moving water in the soil. This study indicates that further work on improving the accuracy of the model with respect to representation of plant transpiration, movement of water in irrigation furrows, and the role of the groundwater in influencing salt concentrations above the water table.

Technical Abstract: Cl- transport in the vadose zone was studied for 45 locations in a field in the San Joaquin Valley through a combination of extensive soil sampling and numerical modeling using the Unsatchem 1-D multicomponent solute transport model. Initial soil sampling in November, 1995, generated a data set that defined the initial conditions and soil hydraulic properties required by the model. Four more soil sampling periods, ending in November 1997, provided an extensive data set. Chloride profiles for different locations exhibited a variety of shapes including positive and negative ramps, maxima and sigmoidal shapes. The standard Unsatchem variably-saturated model for uniform water flow (UFM) conditions predicted greater Cl- leaching than occurred. Potential transpiration was increased by a factor of 1.5 from estimated potential transpiration based on data from California Irrigation Management Information System weather station. This increased upward water flow and improved predictions of Cl- profiles. Unsatchem was upgraded to include the effects of immobile water on solute transport. The resulting mobile-immobile transport model (MIM) demonstrated that higher immobile water content and lower mass transfer coefficients further improved the predictions of the Cl- profiles. While MIM model predictions were an improvement on UFM predictions, further improvements may be possible by increased knowledge of the lower boundary condition, the spatial variability in water flow and improved estimation of soil surface evaporation and transpiration rates. Further work is recommended on process modeling to enable prediction of near surface Cl- maxima that proved especially difficult to predict with either model.