Investigation of L Band PALSAR Synthetic Aperture Radar (SAR) Data in Land Cover Mapping

Mapping of wetlands using remote sensing has faced challenges due to different land cover classes and similar reflectance across land cover classes. Synthetic aperture radar (SAR) data have a special ability to detect phenomena because of their penetration capacity, as well as their independence from time and weather conditionsIn this study, radar data were used to map the land cover of Hamoun wetlands in both wet and dry conditions. For this purpose, L-band capacity has been used to investigate land cover classes. Statistical indicators were used for assessing the separation between land cover classesThe images were classified using the method of the support vector machine. The results of the present study show the capacity of the L-band to separate different land cover classes. The results of the accuracy assessment show an overall accuracy of over 80% in preparing the land cover map of the region. Furthermore, because of the penetration of the L-band, it is possible to detect water in the underlying layer of plants. The results of this study showed that the difference between the reflectance of vegetation that their smaller portion is below the water, and vegetation, most of which is above the water is detectable by the L-band. The difference between their backscatter is 6 dB in the HV and about 2 dB in the HH polarization, which indicates the capaity of the HV polarization to separate these two classes.