Exploring the utility of the Enhanced Vegetation Index as rainfall and agricultural proxy in a Caribbean case study event

Abstract

Highly fragile small island states experience disproportionate climate impacts given their limited capacity to implement cost‐effective tools for detecting emerging signals of drying conditions and monitoring systems for sensitive sectors such as agriculture, especially for uncertain, ‘creeping’ events such as droughts. Despite the existence of open‐source Google Earth Engine datasets, untapped potential remains for their full deployment in disaster management infrastructure. Given this gap, this paper explores the utility of the Enhanced Vegetation Index (EVI) for detecting spatio‐temporal variations of Mid‐Summer Drought (MSD) impacts on vegetation in the small island of Jamaica, with emphasis on major historical drought events. Geospatial analyses of EVI datasets from the Terra Moderate‐Resolution Imaging Spectroradiometer (MODIS) between 2000−2015 archived by the United States Geological Survey (USGS), were computed, and validated by station‐based precipitation and production data for selected parishes for historical case study MSD events. Results revealed highly asymmetrical drought impacts, with Jamaica's agriculturally intense Southern coastline displaying the most stressed vegetation (EVI < 0.5). North‐Western and North‐Eastern regions had the healthiest vegetation during the MSD (EVI > 0.6). A ‘fair’ to ‘moderate’ concurrent correlation was found between EVI and precipitation (R > 0.6), with lower correlations vis‐a‐vis agricultural production (R = 0.2–0.4). The results provide evidence of EVI's utility as a drought monitoring tool in a small island context.