A New Algorithm for Soil Moisture Retrieval With L-Band Radiometer

Abstract

A soil moisture retrieval algorithm for bare surface using the L-band radiometer dual-polarization measurements was developed in this study. The newly developed algorithm is based on a simple semi-empirical model, the Hp model. Through analyzing the simulated database of the Advanced Integral Equation Model (AIEM) under SMAP (Soil Moisture Active and Passive) sensor configurations, it was found that the surface roughness can be cancelled out using the relationship between the V and H polarization surface reflectivity. This leads to a new method for estimation of the soil dielectric properties directly from the radiometer measurements. The simulated data and ground radiometer measurements of bare surface obtained from Beltsville Agriculture Research Center (BARC) during 1979-1981 were used for validation. The inversion accuracy with RMSE (root-mean-square error) for simulated data and bare surface experimental data are 0.014 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> /m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> and 0.036 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> /m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> , respectively. The soil moisture retrieval in vegetated area was performed using the PALS radiometer measurements at L-band during Soil Moisture Experiments in 2002 (SMEX02). The retrieval error was higher for the SMEX02 data, with overall RMSE of 0.043 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> /m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> , due to the high vegetation water content. Using SMOS data in Yanco region soil moisture was retrieved and evaluated, the results are encouraging, with the RMSE of 0.043 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> /m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> and 0.041 m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> /m <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> for ascending and descending, respectively. However, the algorithm performed satisfactorily over the all surface cover conditions including bare and vegetated region.