Submitted to: Journal of Hydrometeorology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: July 1, 2008
Publication Date: March 1, 2009
Citation: Crow, W.T., Huffman, G.J., Bindlish, R., Jackson, T.J. 2009. Improving satellite-based rainfall estimates over land using spaceborne surface soil moisture retrievals. Journal of Hydrometeorology. 10(1):199-212. Interpretive Summary: In order to maximize the value of remote sensing observations for agricultural applications we need to develop effective techniques for combining information gleaned from different types of remote sensing measurement types. This paper presents the first viable method for merging remotely-sensed surface soil moisture information with remote sensing estimates of rainfall to improve our ability to globally monitor rainfall accumulations. Such accumulations are required to globally evaluate the onset and evolution of agricultural drought conditions and enhance our ability to detect changes in the global hydrological cycle due to anthropogenic climate change. This research also develops a key new application for remotely-sensed soil moisture retrieval techniques pioneered by ARS scientists and contributes to a long-term effect by a number of ARS scientists to promote a dedicated US soil moisture satellite mission.
Technical Abstract: Over land, remotely-sensed surface soil moisture and precipitation accumulation retrievals contain complementary information that can be exploited for the mutual benefit of both products. Here a Kalman filtering based tool is developed that utilizes a time series of spaceborne surface soil moisture retrievals to enhance short-term (2- to 10-day) satellite-based rainfall accumulation products. Using the National Center for Environmental Prediction's Climate Prediction Center rain gauge data as a validation source, and a soil moisture product derived from the Advanced Microwave Scanning Radiometer aboard the NASA Aqua satellite, the approach is evaluated over the contiguous United States and Mexico. Results demonstrate that, for areas of suitable land cover, the procedure is capable of improving a variety of satellite-based land precipitation products. Special emphasis is placed on demonstrating that the approach can be parameterized in data-poor areas lacking ground-based observations and/or long-term satellite data records.