Soil moisture inversion algorithms using ERS-1, JERS-1, and ALMAX SAR data

Although the potential of microwave radar sensors to map soil moisture has been well-understood, continuous temporal and spatial monitoring of this important hydrological parameter has been limited due to the non-availability of operational satellite-borne sensor systems. However, the recent launch of ERS-1, JERS-1 and ALMAZ satellites carrying on-board SAR instruments has made possible synoptic soil moisture monitoring a reality. These systems operate over a wide range of frequencies, look angles and transmit-receive polarizations, and thus show synergistic advantages when combined for estimating soil moisture. We have developed a neural-network based soil moisture inversion algorithm that uses as inputs radar backscattering data solely from the above sensor systems, and tested the same using simulated data with speckle added. It appears that the neural-network approach yields superior results in mapping moisture patterns compared to the linear statistical inversion technique, although both show comparable errors in volumetric soil moisture estimation.