Surface Soil Moisture Temporal Persistence and Stability in a Semi-Arid Watershed

Abstract

Satellite soil moisture products, such as those derived from the advanced microwave scanning radiometer - EOS (AMSR-E), require calibration and validation in diverse landscape and land cover conditions. Semi-arid regions present a particular challenge because of the high spatial variability and temporal change in surface moisture conditions. This study will address this problem by using data from a soil moisture observing network in semi-arid watershed and temporal stability (persistence) analysis to quantify its' ability to represent the larger (satellite) scale soil moisture average. The watershed utilized, the Walnut Gulch Experimental Watershed (WGEW), has a dense soil moisture sensor network (SMSN) of 19 soil moisture sensors installed at 5 cm depth, distributed over a 150 km2 study region. A study period of 3.5 years was available for this study. In conjunction with this monitoring network, intensive gravimetric soil moisture sampling was conducted as part of the Soil Moisture Experiment in 2004 SMEX04 to the calibrate the network for large-scale. Temporal stability analysis considers how each individual sensor relates to the overall average of the watershed. Stable sensors (and collectively stable networks) are useful for the long-term study of satellite remote sensing products because of their proven reliability and accuracy. The results demonstrate that the WGEW SMSN is an accurate and stable estimator of the watershed average. The root mean square error (RMSE) of the network to the average from the high density SMEX04 sampling is less than 0.01 m3/m3. Future studies should focus specifically on how best to obtain more reliable satellite soil moisture estimates using this soil moisture network.