Remote monitoring of forest insect defoliation.A review.

Abstract

Aim of study: This paper reviews the global research during the last 6 years (2007-2012) on the state, trends and potential of remote sensing for detecting, mapping and monitoring forest defoliation caused by insects. Area of study: The review covers research carried out within different countries in Europe and America. Main results: A nation or region wide monitoring system should be scaled in two levels, one using time-series with moderate to coarse resolutions, and the other with fine or high resolution. Thus, MODIS data is increasingly used for early warning detection, whereas Landsat data is predominant in defoliation damage research. Furthermore, ALS data currently stands as the more promising option for operative detection of defoliation. Vegetation indices based on infrared-medium/near-infrared ratios and on moisture content indicators are of great potential for mapping insect pest defoliation, although NDVI is the most widely used and tested. Research highlights: Among most promising methods for insect defoliation monitoring are Spectral Mixture Analysis, best suited for detection due to its sub-pixel recognition enhancing multispectral data, and use of logistic models as function of vegetation index change between two dates, recommended for predicting defoliation.