Evaluation of the Performance of Time-Series Sentinel-1 Data for Discriminating Rock Units

The potential use of time-series Sentinel-1 synthetic aperture radar (SAR) data for rock unit discrimination has never been explored in previous studies. Here, we employed to discriminate Dananhu formation, Xinjiang group, Granite, Wusu Xishanyao and Diorite Xinjiang, China. Firstly, the temporal variation backscatter metrics (backscatter coefficient coherence) from April October derived Sentinel-1, was analyzed. Then, significant differences SAR among different units were checked using Kruskal–Wallis rank sum test Tukey’s honest difference test. Finally, random forest models used units. As input features, there four groups: (1) metrics, (2) single-date (3) at VV, (4) VH channel. In each feature three sub-groups: coefficient, coherence, combined coherence. Our results showed that could improve accuracy by roughly 9% (from 55.4% 64.4%), compared data. Both VV polarization provided comparable results. Coherence complements when discriminating Among six units, Granite group can be better differentiated than other Though result still leaves space improvement, this study further demonstrates great discrimination.