Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: February 1, 2012
Publication Date: February 5, 2012
Citation: Parinussa, R.M., Holmes, T.R., Crow, W.T., De Jeu, R. 2012. Large scale evaluation of soil moisture retrievals from passive microwave observations [abstract]. Meeting Abstract. 2012 CDROM. Technical Abstract: For several years passive microwave observations have been used to retrieve surface soil moisture from the Earth’s surface. Several satellite sensors such as the Advanced Microwave Scanning Radiometer-EOS (AMSR-E) and WindSat have been used for this purpose using multi-channel observations. Large scale validation of these retrievals is generally hampered by a lack of ground-based observation networks with sufficient spatial density to be accurately up-scaled to the resolution of satellite-based soil moisture retrievals. In response to this challenge, two new global evaluation techniques have been proposed which circumvent the need for extensive ground-based soil moisture observations. The first technique (Rvalue) is based on calculating the correlation coefficient between known rainfall errors and Kalman filter analysis increments realized during the assimilation of remotely sensed soil moisture into an antecedent precipitation index. The second technique is based on a so-called Triple Collocation (TC) analysis, which is a statistical tool for estimating the root mean square error (RMSE) of a set of three linearly related data sources with independent error structures. These two newly-developed, large-scale soil moisture evaluation techniques are applied for cross-verification on a global scale. Both techniques are also used to determine the sensitivity of soil moisture retrievals to land surface temperature estimates by artificially degrading the satellite signal used for the retrieval of this important parameter. Instead of coincident land surface temperature observations from the same satellite, external sources for land surface temperature are also evaluated using the same evaluation techniques. Finally, both day- and night-time observations are evaluated separately to determine the impact of the different physical conditions during day- and night-time. The evaluation results produced by the Rvalue and TC soil moisture verification approaches show a high mutual consistency (R2 = 0.95), which lends confidence to their interpretation as robust evaluation techniques. They show that the quality of soil moisture retrievals has a strong link with density of the vegetation cover of the observed area. This link was also found when evaluating the different scenarios for the land surface temperature input and when comparing soil moisture retrievals from day- and night-time observations. This study could be used as a framework to evaluate retrievals from the recent Soil Moisture and Ocean Salinity (SMOS) and future Soil Moisture Active and Passive (SMAP) missions. USDA is an equal opportunity provider and employer.