The focus of this study is on the role of precipitation uncertainty in the estimation of soil texture and soil hydraulic properties for application to land–atmosphere modeling systems. This work extends a recent study by Santanello et al. (2007) in which it was shown that soil texture and related physical parameters may be estimated using a combination of multi-temporal microwave remote sensing, land surface modeling, and parameter estimation methods. As in the previous study, the NASA-GSFC Land Information System modeling framework, including the community Noah land surface model constrained with pedotransfer functions (PTFs) for use with the Parameter Estimation Tool, is applied to several sites in the Walnut Gulch Experimental Watershed (WGEW) in southeastern Arizona, USA, during the Monsoon’90 experiment period.

It is demonstrated that the application of PTF constraints in the estimation process for hydraulic parameters provides accuracy similar to direct hydrologic parameter estimation, with the additional benefit of simultaneously-estimated soil texture. Precipitation uncertainty is then represented with systematically varying sources, from the high-density precipitation gauge network in WGEW to lower-quality sources, including spatially averaged precipitation, single gauges in and near the watershed, and results from the continental-scale North American Regional Reanalysis dataset. It is demonstrated that the quality of the input precipitation dataset, and particularly the accuracy of the dataset in both detection of convective (heavy) rainfall events, and in reproduction of the observed rainfall rate probabilities, is a critical determinant in the use of successive remote sensing results in order to establish and refine estimates of soil texture and hydraulic properties.