Analysis of polarimetric coherence data and derived roll-variant parameters of Sentinel-1 data for discriminating mangrove types in Pichavaram wetlands

Abstract

The major focus of the present study is the polarimetric coherence, which corresponds to the correlation between two polarized channels of Synthetic Aperture Radar (SAR) datasets. The capability of the polarimetric coherence to detect target scatters utilizing the entire rotation domain rather than the fixed radar line of sight is the fundamental concept applied in this study. The present study attempts to test the potential of these concepts in mangrove types discrimination using the Sentinel-1 dual polarimetric (VV and VH) dataset. The study targets to characterize the Rhizophora and Avicennia mangrove types of the Pichavaram region of the eastern coast of India. The roll variant parameters were estimated using the polarimetric coherence values derived in the rotation domain for non-redundant angles. The seasonal influence on these parameters was prominent and confirmed through wet and dry season analysis using the ombrothermic diagram, which depicts the temperature and rainfall data. The temporal analysis with a monthly frequency confined to the period of 2018 to 2020 and revealed the capability of discriminating the selected mangrove types better in the pre-monsoon period. The results also unveiled the sensitivity of the polarimetric coherence and associated parameters to the abrupt change in climatic factors, especially during the dry to wet season transition. Finally, the results are verified with a statistical analysis performed using Coefficient of Variation (CV) and Scale Factor. The scope of the present study can be extended to the utility of roll variant parameters along with polarimetric decomposition derived parameters for better characterization of target scatterers. KEY POLICY HIGHLIGHTS Polarimetric coherence at a fixed radar line of sight can be extended over a rotation domain to discriminate mangrove types. The seasonal influences can be captured through polarimetric coherence pattern analysis. The combined climate and polarimetric coherence pattern analysis identify the appropriate period for mangrove type discrimination.