Satellite Image-Based Time Series Observations of Vegetation Response to Hurricane Irma in the Lower Florida Keys

Abstract

High-resolution satellite imaging represents a potentially effective technique to monitor cyclone-caused environmental damage and recovery over large areas at a high spatial scale. This study utilized a 10-m resolution Sentinel satellite image series to document vegetation changes in a portion of the Florida Keys, USA, over which the core of Category 4 Hurricane Irma passed on 10 September 2017. A previously assembled field survey was used to establish land-cover patterns in the satellite data, and concurrent field measurements verified post-hurricane changes. Normalized difference vegetation index (NDVI) was utilized as a tracer for pre-storm baseline patterns and through 19 post-storm months. NDVI patterns show that the severity of vegetation damage varied appreciably across the area, with the least damage on islands in the western sector of the hurricane’s eye and around its center, and greatest damage on islands just east of the eye. The data reveal that for 2.5 months after the storm, multiple inland vegetation classes showed substantial early regrowth. However, mangrove forests were more negatively affected. The storm caused extensive mortality of black mangrove (Avicennia germinans) and red mangrove (Rhizophora mangle), corresponding to more than 40% of the total mangrove area on some islands. The full extent of mangrove die-off was not immediately evident, and increased progressively through the first few months after the storm. In addition to demonstrating the utility of high-resolution satellite image series for post-hurricane environmental assessment, this study reveals high-resolution links between vegetation types, their location within the cyclone, and the extent of post-storm recovery.