Effect of dielectric properties of moist salinized soils on backscattering coefficients extracted from RADARSAT image

Abstract

This paper presents the experimental results regarding changes in the dielectric properties of artificially moistened and salinized soils and on soil samples taken from a salt lake. The complex dielectric constants of soil samples were measured using a microwave network analyzer. We evaluated the real and imaginary parts of dielectric constants of artificially moistened and salinized soil samples prepared in the laboratory as a function of microwave frequency, salinity, and water content. The soil moisture strongly affected the real part of dielectric constant /spl epsiv/'. The frequency and the salinity of soils have little influence on the real part of dielectric constant /spl epsiv/'. The imaginary part of the dielectric constant /spl epsiv/'' is strongly affected by both the salinity and moisture of soil samples, especially at low frequencies. When f<2 GHz, the imaginary part /spl epsiv/'' decreases with each increment of frequency; in contrast, at higher frequencies, /spl epsiv/'' tends toward a constant. The results show that in the frequency range of 1-6 GHz, the imaginary part /spl epsiv/'' has greater sensitivity to soil salinity. Therefore, C and L band are more suitable for soil salinity detection, but L band is better than C band. The dielectric constant measurements for soil samples collected in Jilantai Salt Lake are in agreement with the results of the artificially moistened and salinized laboratory soil samples. These dielectric measurements were subsequently compared with the /spl sigma//sup 0/ backscattering coefficients extracted from a RADARSAT image (C-HH) that was acquired at same time with the soil sampling at the Jilantai Salt Lake area. We discovered that the correlation coefficient between /spl sigma//sup 0/ extracted from the RADARSAT image and /spl epsiv/ measured in the soil sample is 0.70, whereas the correlation coefficient of /spl sigma//sup 0/ and /spl epsiv/' is 0.27. The correlation coefficient between /spl sigma//sup 0/ and the soil sample salinity is 0.69. This suggests that soil salinity has a significant contribution to the backscattering coefficient /spl sigma//sup 0/ recorded in a synthetic aperture radar (SAR) image. Consequently, a SAR image can be a useful tool for monitoring soil salinity.