Abstract
Since detection in 2002, emerald ash borer (EAB), Agrilus planipennis, has caused
widespread mortality of ash trees (Fraxinus spp.) trees across the eastern and midwestern United
States. In this study, I attempted to map EAB-driven ash mortality in New York State based on
analysis of time-series Landsat imagery. Decreasing trends in spectral indices of canopy
greenness were weaker and less consistent than expected in stands with known ash mortality. I
found that time-series NBR and TCG indices, segmented and smoothed with the Landtrendr
algorithm, provided a limited basis for broad-scale ash mortality detection. The highest
producer’s accuracy among map outputs was 58%, although these maps likely exhibited a
tradeoff with user’s accuracy.