Skip to main content
ARS Home » Research » Publications at this Location » Publication #242334
Title: Estimating effective soil properties of heterogeneous areas for modeling infiltration and redistribution

Author
item Ahuja, Lajpat
item Ma, Liwang
item Green, Timothy

Submitted to: Soil Science Society of America Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/3/2010
Publication Date: 5/1/2010
Citation: Ahuja, L.R., Ma, L., Green, T.R. 2010. Estimating effective soil properties of heterogeneous areas for modeling infiltration and redistribution. Soil Science Society of America Journal. 74(5):1469-1482.

Interpretive Summary: Field scale water infiltration and soil-water models require spatially-averaged “effective” soil hydraulic parameters to represent the average flux and storage. The values of these effective parameters vary for different conditions, processes, and component soils in a field. We investigated whether: (1) an effective field saturated hydraulic conductivity, Ks,eff, and correlated hydraulic parameters derived from instantaneous ponded-water infiltration for certain conditions could give reasonable results for infiltration of lower rainfall rates; and (2) these effective parameters could also give acceptable average results for soil water redistribution. The results of this “Effective Parameter Set 1” (EPS-1) were compared with those of another parameter set, EPS-2, where the above Ks,eff was combined with other hydraulic parameters that were arithmetic mean values of the component soils in the field, and with EPS-3 where all parameters were arithmetic mean values of the component soils. The RZWQM2 model was used to explore these objectives for five different cases of a heterogeneous field, comparing results from effective properties with weighted mean values of the components of the fields at four different rainfall intensities. The EPS-1 gave the best results for infiltration cases, but failed to give good results for the weighted mean soil water content distributions in the fields 7 days after infiltration. The EPS-2 seemed to be a reasonable compromise for infiltration and soil water contents for all rainfall intensities. The results are an important step toward scaling-up or estimating effective properties of heterogeneous field or hydrologic units in watershed modeling.

Technical Abstract: Field scale water infiltration and soil-water and solute transport models require spatially-averaged “effective” soil hydraulic parameters to represent the average flux and storage. The values of these effective parameters vary for different conditions, processes, and component soils in a field. For practical applications, however, the effective properties may closely approximate the averaged results for a certain range of conditions. The objectives of this study were to investigate whether: (1) an effective field saturated hydraulic conductivity, Ks,eff, and correlated hydraulic parameters derived from instantaneous ponded-water infiltration for certain conditions could give reasonable results for infiltration of lower rainfall rates; and (2) these effective parameters could also give acceptable average results for soil water redistribution. The results of this “Effective Parameter Set 1” (EPS-1) were compared with those of another parameter set, EPS-2, where the above Ks,eff was combined with other hydraulic parameters that were arithmetic mean values of the component soils in the field, and with EPS-3 where all parameters were arithmetic mean values of the component soils. The RZWQM2 model was used to explore these objectives for five different cases of a heterogeneous field, comparing results from effective properties with weighted mean values of the components of the fields at four different rainfall intensities: 100 (incipient ponding), 10, 5, and 2.5 cm/h. The mean absolute error for infiltration over four durations (0.5, 1, 2, 4 h) generally increased with decreasing rain intensity, but not consistently for all cases. The errors were generally the least with EPS-1 and most with EPS-3. For a field containing three uniform soils (sandy loam, silt loam, silty clay loam) in equal proportion, the mean error from EPS-1varied from 2.1% for the instantaneous ponding to 16.2% for 2.5 cm/h rainfall; the corresponding values for EPS-3 were 19% and 25.3%, respectively. Lowest errors were obtained with a single spatially variable sandy loam soil in a field (EPS-1 2.4-4.7%; EPS-3 2.6-11.5%), and nearly the highest with all three spatially variable soils (EPS-1 3.6-24%; EPS-3 34-44.2%). However, the EPS-1 failed to give good results for the weighted mean soil water content distributions in the fields 7 days after infiltration. With EPS-2, overall results were slightly worse for infiltration but much better for soil water content distributions. The EPS-3 gave the worst overall results. The EPS-2 seemed to be a reasonable compromise for infiltration and soil water contents for all rainfall intensities.