Unveiling Forest Resilience Changes in Response to Insect Disturbance: A Comprehensive Analysis Using PlanetScope Time Series

Abstract

The outbreaks of insect disturbance, especially spongy moth defoliations, have had significant impacts on the overall health and vigor of New England's forests. However, it is difficult to use ground-based methods to continuously monitor forest disturbances and estimate their impacts on ecosystem resilience over broad areas. To solve this problem, we propose to develop a system that integrates insect disturbance detection and resilience monitoring using daily PlanetScope imagery. We will conduct the COntinuous monitoring of Land Disturbance (COLD) algorithm and random forest classification to map various forest disturbance types, including spongy moth outbreaks. The forest resilience will be represented by the lag-1 temporal autocorrelation (TAC) of the Enhanced Vegetation Index (EVI) time series. We will select TAC thresholds to categorize forest health into three main classes: healthy, declining, and deceased. This research aims to establish a high-resolution database on forest disturbance and condition shifts. This database has the potential to identify hazardous trees near roadside utilities and offer guidance for vegetation management.