Soil moisture sensing by L-band radiometry for prairie grassland

A Land Surface Process/Radiobrightness (LSP/R) model is integrated with a dynamic learning neural network (DLNN) to examine the impact of L-band on radiometric sensing of soil moisture for prairie grassland. The LSP/R model predictions of brightness temperature, and surface parameters are used as the training and evaluating data for the DLNN. Both horizontally- and vertically-polarized brightnesses at 1.4, 19, and 37 GHz for an incidence angle of 53 degrees make up the input nodes of the DLNN. The corresponding output nodes compose of land surface parameters, canopy temperature and water content, and soil temperature and moisture (uppermost 5 mm). Under no noise conditions, the root mean square (rms) errors between the retrieved surface parameters and the reference are smaller than 2% for a four-channel case with 19 and 37 GHz brightnesses as the inputs of the DLNN. The rms errors are reduced to within 0.5% if additional 1.4 GHz brightnesses are used (a six-channel case). The results demonstrate that 1.4 GHz is a better frequency in probing soil parameters than 19 and 37 GHz. In addition, the proposed inversion approach on the radiometric sensing of the land surface parameters is promising