Location: Great Basin Rangelands ResearchTitle: Process-based modeling of infiltration, soil loss and dissolved solids on saline and sodic soils
|NOUWAKPO, SAYJRO - University Of Nevada|
|ARSLAND, AWADIS - University Of Nevada|
|GREEN, COLLEEN - Bureau Of Land Management|
|AL-HAMDAN, OSAMA - Texas A&M University|
Submitted to: Transactions of the ASABE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/1/2017
Publication Date: 2/1/2018
Citation: Nouwakpo, S.K., Weltz, M.A., Arsland, A., Green, C.H., Al-Hamdan, O. 2018. Process-based modeling of infiltration, soil loss and dissolved solids on saline and sodic soils. Transactions of the ASABE. 61, 1033– 1048. https://doi.org/10.13031/trans.12705.
Interpretive Summary: Not applicable
Technical Abstract: The Colorado River is a central resource of the Western United States but is vulnerable to excessive salt load. To improve knowledge on surface processes controlling salt loading, a series of rainfall simulation experiments were conducted in saline rangelands of the Upper Colorado River Basin (UCRB). In this paper, data from these rainfall simulation experiments were used to develop predictive equation for the process-based Rangeland Hydrology and Erosion Model (RHEM). Runoff and soil loss prediction performances were assessed with the Nash-Sutcliff Efficiency (NSE), the coefficient of determination (R2) and the percent bias (PBIAS). Calibration on 36 individual plots, randomly selected to cover each treatment, yielded improved runoff prediction (NSE = 0.73, R2 = 0.74 and PBIAS = 6.93%) compared to the non-calibrated RHEM parameter estimation equation (NSE = 0.65, R2 = 0.68 and PBIAS = 32.03%) when a refined ground cover coefficient was used to estimate the effective hydraulic conductivity Ke. Soil loss prediction on the calibration data was also improved compared to the non-calibrated parameter estimation equation (NSE = 0.94, R2 = 0.94 and PBIAS = 4.25% vs. NSE = 0.81, R2 = 0.85 and PBIAS = 6.47% ) when soil Sodium Adsorption Ration (SAR) was included in the estimation of the splash and sheet erosion parameter (Kss). Improvements in runoff and soil loss predictions on the calibration data were maintained on an independent set of 36 plots from the original rainfall simulation dataset not used for calibration. Overall, soil sodicity was an important consideration in the performance on the newly developed Kss parameterization equation in this paper. Performance on sodic soils (SAR = 15) gained the most from the inclusion of SAR in the Kss estimation. Salt load was linearly related to soil loss (R2 = 0.94) and this linear model performed well to RHEM soil loss predictions into salt load estimates. These new developments will provide a physically-based modeling scheme to land managers for predicting rainfall-driven soil and salt load to surface waters of the UCRB.