Monitoring Forest Change in Southeast Asia: Case Studies
This joint study by the US Forest Service’s Remote Sensing Application Centre and USAID LEAF evaluated three automated forest loss detection algorithms in four USAID LEAF landscapes using Landsat data spanning 2001 to 2013. Google Earth Engine (GEE) was used as an efficient cloud-based computation platform providing massive computing infrastructure available to all Asian countries with limited information technology infrastructure.
The evaluated algorithms were CLASlite, Hansen’s Global Forest Loss product, and Multiple Linear Trend Analysis (MLTA). A quantitative accuracy assessment was also conducted using the Timesync Landsat visualization tool across a total of 2,000, 30 by 30 meter sample pixels. Results indicate that the Hansen product, while only identifying deforestation, overlaps with much of what the accuracy assessment characterized as forest degradation, thus combining the two in a single class. The CLASlite products generally had the lowest accuracies. The MLTA product had high accuracies in some areas, which indicated that with better calibration the method could potentially meet monitoring needs.