Urban Forest Cover and LST Change Monitoring Through Optical and Thermal Remote Sensing Approach

Integrated monitoring of urban forest cover and land surface temperature (LST) can serve as a baseline for urban planners and legislators to assimilate multi-temporal data to interpret the status and trend of reducing carbon emissions from the loss of urban forest cover over time. With the spectral and thermal information that the optical and thermal remote sensing data offer, the forest cover of Baguio City, Philippines, in 2010 and 2020 using Landsat-5 and Sentinel-2, respectively, was monitored through random forest technique and the LST during the cold and hot seasons was estimated using multi-temporal Landsat-5 and Landsat-8 imageries through a single-channel algorithm. The atmospheric and illumination effects on the surface reflectance of the Landsat and Sentinel images were corrected using Fast Line-of-sight Atmospheric Analysis of Hypercubes and Atmospheric and Topographic Correction methods, respectively. The pansharpening technique was also performed for the Sentinel image and resampling for the Landsat images. The result showed that the forest cover and built-up areas classification in 2010 had an overall accuracy of 0.95 and a kappa coefficient of 0.90, while in 2020 had 0.98 and 0.97. A forest cover loss rate of 0.34