Effects of changing environmental conditions on synthetic aperture radar backscattering coefficient, scattering mechanisms, and class separability in a forest area

Environmental conditions have considerable effects on synthetic aperture radar (SAR) imagery. Therefore, assessing these effects is important for obtaining accurate and reliable results. In this study, three series of RADARSAT-2 SAR images were evaluated. In each of these series, the sensor configuration was fixed, but the environmental conditions differed. The effects of variable environmental conditions were also investigated on co- and cross-polarized backscattering coefficients, Freeman-Durden scattering contributions, and the pedestal height in different classes of a forest area in Ottawa, Ontario. It was observed that the backscattering coefficient of wet snow was up to 2 dB more than that of dry snow. The absence of snow also caused a decrease of up to 3 dB in the surface scattering of ground and up to 5 dB in that of trees. In addition, the backscatter coefficients of ground vegetation, hardwood species, and softwood species were more similar at temperatures below 0 degrees C than those at temperatures above 0 degrees C. Moreover, the pedestal height was generally greater at temperatures above 0 degrees C than at temperatures below 0 degrees C. Finally, the highest class separability was observed when the temperature was at or above 0 degrees C and there was no snow on the ground or trees. (C) 2017 Society of Photo-Optical Instrumentation Engineers (SPIE)