Comparison of two retrieval methods with combined passive and active microwave remote sensing observations for soil moisture

The brightness temperature (BT) and backscattering coefficient (BSC) measured simultaneously by passive and active microwave sensors have great potential for the estimation of land surface soil moisture (SM). Several methods with combined passive and active microwave remote sensing observations for SM have been reported. Usually, the use of these methods, requires an accurate roughness condition, especially when dealing over the bare surface or surface with low vegetation. In this paper, two different retrieval methods for estimating SM using synthetic microwave remote sensing data were compared. The difference between the twomethods is theway to estimate the roughness parameter. The threeparameter retrieval approach (THRA) takes the roughness parameter as a variable to be retrieved, and derives land surface parameters (e.g., SM, surface temperature (ST) and roughness) simultaneously from BT and BSC using nonlinear algorithms. The twoparameter retrieval approach (TWRA) retrieves roughness from BSC firstly, then takes the roughness as an input parameter to retrieve SM and ST from BT. The two retrieval methods were applied on AMSR-E (Advance Microwave Scanning Radiometer for EOS) and QuikSCAT/Seawinds (The Seawinds scatterometer on NASA’s (National Aeronautics and Space Administration) Quick Scatterometer) observations which have been carried out for the SMEX03 (Soil Moisture Experiment 2003) region in ON, the north of Oklanoma. The comparison results shows TWRA has achieved a higher accuracy than THRA in dealingwith the active and passive microwave observations at different overpass times. © 2010 Elsevier Ltd. All rights reserved.