Enhanced hydrologic simulation may not improve downscaled soil moisture patterns without improved soil characterization. Issue 3 (22nd April 2020)

• Record Type: Journal Article
• Authors: Pauly, Matthew J.; Niemann, Jeffrey D.; Scalia, Joseph; Green, Timothy R.; Erskine, Robert H.; Jones, Andrew S.; Grazaitis, Peter J.
• Is Part Of: Soil Science Society of America Journal. Volume 84:Issue 3(2020)
• Page Start: 672
• Page End: 689
• Publication Date: 2020-04-22
• Subjects: Soils -- United States -- Periodicals; Soil science -- Periodicals; Periodicals; 631.4973
• DOI: 10.1002/saj2.20052
• Languages: English
• ISSNs: 0361-5995

Abstract

Abstract: Fine‐resolution soil moisture maps (10–30 m grid cells) are useful for many applications, including agricultural production and off‐road vehicle mobility. Fine‐resolution maps of soil moisture can be produced by downscaling coarse‐resolution soil moisture data from remote sensing. Soil moisture downscaling has been conducted by modeling key hydrologic processes in the soil, but current downscaling methods neglect two key aspects of soil hydrology: surface runoff and residual water content. The objective of this study is to understand how these hydrologic considerations affect spatial patterns of soil moisture and whether their inclusion improves downscaling performance. The Equilibrium Moisture from Topography, Vegetation, and Soil (EMT+VS) downscaling model is used to address these questions. Surface runoff is introduced to the model by implementing an infiltration capacity based on the saturated hydraulic conductivity. Residual water content is included by assuming that all hydrologic processes cease when soil moisture is at or below the residual water content. The pre‐existing EMT+VS model is compared with the new models using data from northeastern Colorado. The results show that both model additions introduce more complex dependence on the spatial variability of soil texture, which is difficult to estimate accurately. Including surface runoff provides only a small improvement in model performance, and residual water content shows almost no effect on the performance in this case. Thus, the simplifying assumptions of the pre‐existing model provide an appropriate level of complexity, given the uncertainties of the spatial input data.