Monitoring pine defoliation due to the processionary moth at regional scale from MODIS time series

Insect related damage has severe ecological and socio-economic impacts by causing large-scale mortality of vegetation and thereby affecting biogeochemical cycles and forest industry. The objective of this study was to evaluate the potential use of 250 m MODIS data time-series to monitor damage caused by the pine processionary moth (PPM), which is the most important pine defoliator of the maritime-pine forest in southwest France. We compared the in situ observed defoliation rates (DR) and the multi-seasonal variations of the MODIS-sensor derived MOD13Q1 product. We exploited the winter anomalies in the EVI signals to empirically detect and quantify the PPM defoliation. Different EVI-based anomaly metrics were tested by linear regression over a test site of 18 km². The best correlated relationship (R²=0.908, p<;0.0001) to observed DR was used to define an empirical model. The predictions of DR after model inversion were validated at another site with a larger extent (16 km ×16 km). A robust empirical model based on the winter decrease in EVI-based anomaly metrics was found to be suitable to retrieve the spatial pattern of defoliation at the landscape scale.