Mapping Drought-Resilient And Groundwater-Dependant Vegetation With Sentinel-1 Sar Data

Mapping of vegetation that relies on perennial, groundwater-related water sources helps understand groundwater dynamics and aquifers. Identification of where such vegetation occurs enables its protection if groundwater resources are developed. In this study, we explore how indexes based on C-band radar imagery can be used to identify these ecosystems in remote areas of Australia. The approach assumes that, during dry times, difference in access to water creates a sufficient contrast between these ecosystems and other vegetation population for detection via radar imagery products. Hence radar images at the driest times are selected by analyzing GRACE data. The thresholds applied to radar vegetation indexes are optimized to fit in-situ observations. The results show that InSAR coherence contains the most information on groundwater-dependent vegetation. They also show that the radar vegetation index (RVI) adds performance benefits when used jointly with InSAR coherence.