Estimating the aboveground biomass of a forest-protected area in Bangladesh using Google Earth Engine

Abstract

Abstract Aboveground biomass (AGB) is an important forest biophysical property accurate measurement of which is imperative as it plays a crucial role in understanding global climate change. Among the direct and indirect approaches to measuring AGB, remote sensing based indirect approach has been widely used for research and managerial purposes. In this study, a workflow was developed integrating Google Earth Engine (GEE) to estimate the AGB of Khadimnagar National Park (KNP). Collected field AGB data and ten remote sensing vegetation indices were used as dependent and independent variables respectively in Multiple Linear Regression and the modeled equation was integrated into the GEE to create the estimated map of KNP. According to the different statistical indices, the three best-performing models were not reliable enough to depict the spatial distribution of AGB. Additionally, it was found that Chlorophyll Vegetation Index (CVI), Green Leaf Index (GLI), Atmospherically Resistant Vegetation Index (ARVI), and Normalized Difference Vegetation Index (NDVI) were the most important variables in modeling AGB. However, initially, the predicted AGB information of this work will be useful to understand the vegetational health and condition of an important forest-protected area like KNP. Furthermore, a similar AGB modeling approach can be used to map the other forested areas of Bangladesh with different remote sensing datasets and topographical and climatic datasets.