Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: March 25, 2011
Publication Date: April 11, 2011
Citation: Crow, W.T. 2011. Using scatterometer-based surface soil moisture products to optimally calibrate land data assimilation systems [abstract]. EUMETS/ESA Scatterometer Science Conference. 2011 CDROM. Technical Abstract: Land data assimilation systems are designed to merge uncertain land surface model predictions with error-prone observations. Ingestion into a data assimilation systems represents a critical pathway towards key applications goals for remotely-sensed surface soil moisture products. However, the effectiveness of such systems is highly dependent on the availability of accurate modeling and observations error covariance information. Lacking such information, data assimilation systems will sub-optimally merger remotely-sensed surface moisture information with models and partially squander the value of remotely-sensed data products. Unfortunately, the quality of error information currently fed into land data assimilation systems is typically poor and little progress has been made towards the development of robust adaptive land data assimilation systems (designed to estimate such error information during the operational cycling of a data assimilation system). However, recent progress has been made in this area through the application of so-called “triple collocation” techniques to estimate the observational error covariance of remotely-sensed surface soil moisture data products. Such approaches estimate error magnitudes in a given geophysical variable by averaging across variations within three independently-obtained and coincident estimates of the variable. The availability of a scatterometer-based surface soil moisture product allows such an analysis to be conducted in concert with a radiometer-based soil moisture product and a model-based soil moisture product. Here errors in surface soil moisture retrievals obtained from the Advanced Microwave Scanning Radiometer (AMSRE-E) are estimated via a triple-collocation exercise conducting using European Remote Sensing (ERS-1 and -2) scatterometer soil moisture data and land surface model. Results from this analysis are then used to optimize the performance of a land data assimilation system designed to ingest AMSR-E soil moisture retrievals over three heavily-instrumented watershed-scale (102 to 402 km2) sites in the United States. Validation results based on extensive ground-based observations at these sites reveal that the procedure significantly enhances the quality of surface soil moisture estimates relative to existing adaptive filtering approaches. In this way, the availability of a diversity of remotely-sensed soil moisture products (i.e. both radiometer and scatterometer-based) provides a key opportunity to optimally calibrate a land data assimilation system. As a consequence, even in cases in which they are not directly assimilated, the mere availability of a scatterometer-based soil moisture product can enhance the functioning of a land data assimilation system. The extension of the analysis to ASCAT and NASA Soil Moisture Active/Passive (SMAP) data products will be discussed.