Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #293986
Title: Leveraging simultaneous SMOS and ASCAT soil moisture products for enhanced hydrologic prediction

Author
item Crow, Wade
item CHEN, FAN - Science Systems, Inc

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 6/1/2013
Publication Date: 7/21/2013
Citation: Crow, W.T., Chen, F. 2013. Leveraging simultaneous SMOS and ASCAT soil moisture products for enhanced hydrologic prediction [abstract]. Proceeding of the 2013 Geoscience and Remote Sensing Symposium, July 21-26, 2013, Melbourne, Australia. 2013 CDROM.

Interpretive Summary:

Technical Abstract: Runoff predictions obtained from rainfall runoff model are typically degraded for a wide variety of error sources including the inaccurate specification of pre-storm soil moisture conditions (determining infiltration capacity) and random error in rainfall inputs (especially in areas of a world lacking adequate ground-based rainfall instrumentation). Recent work has demonstrated that remotely-sensed surface soil moisture retrievals can address both of these error sources and (thus) potentially contribute to hydrologic applications relying on accurate stream flow forecasts. This presentation will describe recent efforts to simultaneously leverage passive microwave soil moisture retrievals form the Soil Moisture Ocean Salinity (SMOS) mission with scatterometer retrievals from the Advanced Scatterometer (ASCAT) aboard the METOP-A satellite to enhance rainfall-runoff modeling within a large-number of medium-scale (1000 to 10,000 km2) basins in the Southern United States. The approach will be based on the application of the Soil Moisture Analysis Rainfall Tool (SMART) to correct satellite-based precipitation products using remotely-sensed surface soil moisture retrievals (Crow et al., 2011). In parallel, the same set of soil moisture retrievals will be assimilated into a rainfall-runoff model - here the operational Sacramento (SAC) model applied operationally by the United States National Weather Service - using an Ensemble Kalman filter (EnKF) in order to improve the characterization of pre-storm soil moisture levels. Therefore, when SMART-corrected rainfall is fed into the SCA model, errors associated with both the inaccurate specification of antecedent soil moisture conditions and uncertain rainfall observations will be corrected simultaneously. Past work has demonstrated that such simultaneous correction can be applied without violating any statistical assumptions underlying the application of the EnKF and/or SMART (Crow and Ryu, 2009). SMART soil moisture corrections are based on the application of a simple Antecedent Precipitation Index (API) model forced by a satellite-based rainfall product. Next, remotely-sensed surface soil moisture retrievals are assimilated into the API model using a Kalman filter (KF) at discrete times at which they are available. This assimilation process entails the KF-based definition of analysis increments (i.e., net additions/subtractions of water made to the API state in response to comparisons against SMOS and ASCAT surface soil moisture retrievals). These analysis increments are then correctively applied to the original satellite-based rainfall time series (here the TRMM-based 3B40RT product). Past results have demonstrated that such corrections can compensate for a substantial fraction of random errors present in 3B40RT rainfall accumulation estimates. The simultaneously assimilation of SMOS and ASCAT products into the SAC model is accomplished via a 30-member EnKF implementation. Note that, to avoid cross-correlation between modeling and observation errors, the EnKF analysis cycle is based solely on SAC model forecasts obtained using uncorrected satellite-based rainfall data. SMART-corrected rainfall is only applied in a single post-processing step where the SAC model is initialized using the EnKF soil moisture analysis and then forced using SMART-corrected TRMM 3B40RT rainfall data (Crow and Ryu, 2009). The presentation will show results for applying this approach within a number of medium-scale watersheds within the South-central United States.