Diagnosing pristine pine forest development through pansharpened-surface-reflectance Landsat image derived aboveground biomass productivity

Abstract

A map of natural forest aboveground biomass (AGB) can serve as a baseline for forest managers to integrate multi-temporal information to account for the trend of reducing carbon emission from the forest degradation over time. Taking advantage of spectral and spatial enhancement in Landsat OLI images, a machine learning based kNN algorithm was applied to derive AGB maps of a pristine forest of Pinus kesiya in 2014 and 2019 via allometric models of the pine and inventory data. AGB productivity of the study site was derived and examined for diagnosing the pristine pine forest development. The radiance of the Landsat image was first restored and then normalized by the FLAASH model, a radiation transfer model-based algorithm used to derive surface reflectance image, and finally a pansharpening technique was applied to improve the spatial resolution while retaining the spectral resolution of the reflectance image. Results showed that the pansharpened surface reflectance image was able to derive AGB maps with a satisfactory accuracy of root mean square percentage error, RMSPE = 15%, which shows significant improvement on the performance of AGB estimation using the raw image (RMSPE = 56%) and non-pansharpened surface reflectance image (RMSPE = 48%). The loss of AGB stocks is mainly caused by landslide, forest fire, and agricultural expansion, particularly the use of fire for land preparation for cropping. Land covers other than pine stands have complicated the edge of the pristine pine forest. The non-forest cover types surrounded the forest enlarged the boundary of the forest and made the pristine forest slightly fragmented. The results also showed that the AGB loss/gain of a forest parcel is negatively/positively related to the size of the parcel. Appropriate forest land zoning and community-based forest management for the surrounding non-pristine pine forest could have the benefit of mitigating the degradation of the pristine pine forest and improving the biomass stocks for better REDD achievements.