Evaluating biochemical and biophysical characteristics of tropical deciduous forests of eastern India using remote sensing and in-situ parameter estimation

Abstract

Owing to such a crucial role that forests perform in an ecosystem, studying the vegetation dynamics in terms of its productivity becomes inevitable. The study aimed to assess forest productivity and biochemical variability of Dalma Wildlife Sanctuary (DWS), Eastern India coupling remote sensing and in-situ measurements of 51 samples comprising 17 different species. Vegetative Photosynthesis Model (VPM) was incorporated for measuring the gross primary productivity (GPP) using Sentinel-2A data. The result showed that microclimatic conditions that determine the nutrient content of the plant, as some plant species(Shorea robusta) present at different locations showed different concentrations of chlorophyll(25.4 mgg-1), N(2.2%), P(0.5%), and K(2.0%). The relationship between biophysical and biochemical properties of the forest was analyzed by coupling laboratory-based leaf samples analysis and satellite-based retrieval methods. High values of GPP was present in the dense forested (core) region dominated by species like Shorea robusta, Bauhinia tomentosa, Butea monosperma, Syzygium cumini, Cleistanthus collinus, Diospyros melanoxylon, etc. as compared to buffer regions. The central region of DWS dominated by dense vegetation contributed higher GPP value of 21 gCm-2day-1 to 24 gCm-2day-1. The degraded peripheral area exhibited GPP value of 8 gCm-2day-1. Shorea robusta being the most dominant species in this region, exhibited total chlorophyll of 25.4 mgg-1 (lab estimates) with a standard error of 2.0 mgg-1. Correlation analysis between Sentinel-2 satellite-based biophysical parameters and laboratory-based biochemical parameters showed a close relationship between the variables. The correlation between GPP and FAPAR was found to be more profound (R2 = 0.625), whereas GPP and LAI showed a moderate correlation of R2 = 0.58. Nitrogen and Phosphorous content in leaf samples showed a negative relationship (R2 = 0.414) indicating their individual concentration do not actuate to the coefficient of correlation. With rational limitations, the study finds wide application for studying carbon cycle and biochemical assessment in the region with remote sensing driven techniques backed by in-situ parameter estimation.