Detecting long-term global forest change using continuous fields of tree-cover maps from 8-km advanced very high resolution radiometer (AVHRR) data for the years 1982-99

This paper describes global changes in forest cover from 1982 to 1999 based on the 8-km Pathfinder Advanced Very High Resolution Radiometer (AVHRR) data set. The procedure involves the use of a regression tree in predicting percent tree cover for the years 1982-99. Training data are created from high-resolution imagery and used with phenological metrics derived from the annual AVHRR time series. Using the 18 years of estimated tree cover, and based on a thresholding approach, we identified locations where change in tree cover has occurred. The change sites were then compared to a set of high-resolution deforestation analyses to yield area estimates of deforestation and regrowth. Percent tree cover was found to have decreased globally, from the 1980s to 1990s, in contrast to United Nations Food and Agriculture Organization (FAO) reports of a global increase in forest cover. Latin America and tropical Asia are the two dominant deforestation regions. Paraguay shows the highest rate of forest clearing over the time series, while Indonesia had the greatest increase in deforestation from the 1980s to 1990s. We also suggest that the percent tree-cover maps can be used in standardizing national forest statistics, as an objective means of identifying hot spots of change, and for facilitating ecosystem monitoring.