REGIONAL ESTIMATION OF SOIL-MOISTURE USIN G REMOTE-SENSING

Several pilot projects have demonstrated that estimation of soil moist ure over a large area can be done using remote sensing. Three main met hods have been rested with some success: thermal inertia, passive micr owave and synthetic aperture radar (SAR). The advantages and limitatio ns of each approach were summarized. Most Canadian research has focuse d on SAR data. It has shown that several parameters can affect the acc uracy of soil moisture estimation using radar such as incidence angle. roughness, polarization and frequency. The data collected during the SIR-C/X-SAR experiment in Altona, Manitoba, were used to evaluate the impact of incidence angle on soil moisture estimation accuracy. Incide nce angle was the most significant Factor to explain the signal variat ions over lime. The effect of incidence angle (38 degrees to 58 degree s) on the signal was linear in October. Correlation between soil moist ure and the signal was higher with surface (0-2.5 cm) measurements in the wet period (April) but there was no significant correlation during the dry period (October). A statistical model using soil moisture and incidence angle in April showed that an increase of 1 degrees in inci dence angle could decreased the C-HH signal by 0.25 dB and the L-HH si gnal by 0.30 dB. Such variation would generate a change of 2% (C-HH) a nd 5% (L-HH) in soil moisture estimation.