Characterizing Spatiotemporal Patterns of Spongy Moth Defoliation in Upstate New York Using Sentinel-2
Cameron Coles, Cornell University
European spongy moth (formerly European gypsy moth, Lymantria dispar dispar) is an invasive insect in the northeastern United States whose outbreaks can cause extensive defoliation. The consequences of multi-year outbreaks can be seen in all parts of the ecosystem including shifting tree species composition, altered food sources and habitat for other organisms, as well as effects on carbon cycling, soil nutrient dynamics, and more. However, our understanding of forest susceptibility to and the population dynamics of spongy moth outbreaks is limited by a lack of high-quality data that can quantify outbreak-driven defoliation across space and time. We used Sentinel-2 MSI data to create higher spatial resolution defoliation estimates of recent L. dispar outbreaks in upstate New York, USA that occurred in 2021 and 2023. The short backlog of Sentinel-2 Level 2a imagery preempts the usage of previous harmonic trend modeling, hence we developed a new approach that utilizes average linear declines (covering 2019 to 2022, when Level 2a data is available) in enhanced vegetation index (EVI) during the growing season (June through August) modeled by a Theil-Sen regression to identify and quantify extended periods of low EVI indicative of defoliation. Our estimates of defoliation corroborate aerial survey observations and highlight additional unmarked regions of defoliation. Additionally, we perform novel analyses highlighting multi-year outbreaks, variability in outbreak phenology, the size and shape distribution of outbreaks, and dependence of outbreak characteristics on landcover type. These results provide insight into the ecology of L. dispar and how it interacts with its hosts, as well as provide a foundation for accurate incorporation of L. dispar outbreaks into ecological models for theoretical and management purposes.