Assessing the Utility of Historic NASA Imaging Spectroscopy Data to Detect Plant Disease in California Vineyards

Plant disease destroys upwards of 15-20% of the global crop yield, a loss of ~220 billion USD annually. The need to rapidly and accurately detect plant disease is increasingly important to the US grape industry, the nation's highest value fruit crop grown on over one million acres and valued at over $162 billion, as the negative impacts of climate change intensify. Despite the abundance of available spectroscopic data and its potential use as an inexpensive tool for scalable, plant disease detection and mapping, spectroscopy remains drastically underutilized and unexplored. Here, we explore the application of historic satellite and airborne spectroscopic datasets to remote detection of Esca complex, a grapevine trunk disease (GTD) in California vineyards. Esca complex is an ideal model system for plant disease sensing because it is acquired at the nursery stage, and therefore present for the grapevine's entire lifespan, despite only causing visible foliar symptoms at specific growth stages. Symptoms include reduced crop yield, crop quality, and vine life span. High spatial and spectral resolution datasets from NASA's Airborne Visible / Infrared Imaging Spectrometer (AVIRIS NG), NASA's ECOSTRESS, and Planet Imagery were mined in collaboration with industry and academic colleagues to determine the spectral properties of Esca infected vineyards. We applied open source, foliar functional trait models to determine trait variation across the established disease gradient. We find that plant disease mapping using historic spectroscopic acquisitions introduces a non-intrusive and cost-effective approach to disease detection and field management. Our findings support a new utility for the wealth of historic NASA spectroscopic data with direct agricultural applications. Our work supports the future use of spaceborne imaging spectroscopy, including NASA's Surface Biology and Geology (SBG), for global plant disease mapping in agricultural systems.