Modeling soil moisture using passive remote sensing

Soil moisture is an important component of analysis in many Earth science and related disciplines. Information about the entire profile can find a wide range of applications in many disciplines. Hydrological models can simulate soil moisture profiles, but there is usually limited subsurface information to constrain the models. Emerging science and technology to measure soil moisture with remote sensing offers a potential source of additional information from which to constrain soil hydrology models. Because passive remote sensing can provide soil moisture information for a thin surface layer of soil, the question becomes one of how to use this information to improve estimates of the soil moisture profile. This study attempts to shed light on this question by using a simple hydrology model (SHM) and with data collected during a microwave remote sensing experiment in Huntsville, Alabama during July 1996 (Huntsville '96). This study shows how the errors in the estimation of soil moisture increase as the sampling interval of meteorological data increases. The root mean square error (RMSE) from the baseline almost triples (i.e. from 0.0352 to 0.0810), as the sampling interval is increased from 6 hr to 12 hr. These errors can be reduced by half if we periodically update the modeled soil moisture estimates with a method that assimilates microwave remote sensing soil moisture estimates with the SHM soil moisture profile. Thus, this study attempts to extend the soil moisture information on the surface layer by microwave remote sensing to the entire profile using SHM.