Impacts of spatial resolutions and data quality on soil moisture data assimilation



[1] The resolution of near-surface soil moisture imagery derived from passive remote sensing sources is expected to be on the order of 30 km in the foreseeable future. While analogous active remote sensing retrievals may be derived at finer scales (∼100 m), these have been shown to be prone to error and increased uncertainties in the presence of vegetation and/or dominant surface roughness effects. In this study, we resort to an extension of multiscale Kalman filtering (MKF) for assimilation of near-surface soil moisture retrievals derived during the Southern Great Plains Hydrology experiment of 1997 (SGP97) into the Three-Layer Variable Infiltration Capacity (VIC-3L) land surface model. Our first objective is to evaluate impacts of resolution of the passive remotely sensed soil moisture imagery (from 0.78 km to 25 km) on the consistency of the assimilation algorithm and the effectiveness of assimilation with regards to improving the prediction of soil moisture states and energy fluxes from VIC-3L. Motivated by recent evidence suggesting that temporal ratios of near-surface soil moisture retrievals from active remote sensing sources may retain much of the soil moisture signal, we additionally assess the value of synthetically derived fine (0.78 km) scale temporal ratios of near-surface soil moisture when assimilated in conjunction with coarse (25 km) resolution retrievals of near-surface soil water content. Temporal ratios of soil moisture may, in general, be thought of as a noisy data set conveying information on the spatial organization of near-surface soil moisture as opposed to reliable estimates of specific volumetric water content values. We find that the joint assimilation of synthetic temporal ratios of soil moisture and coarse volumetric soil moisture retrievals leads to the recovery of relevant spatial features not captured in the 25-km data set or by VIC-3L.