Primary productivity estimation of forest based on in-situ biophysical parameters and sentinel satellite data using vegetation photosynthesis model in an eastern Indian tropical dry deciduous forest

The present study deals with the estimation of satellite-based gross and net primary productivity ( GPP and NPP ) in combination with in-situ observations in Dalma Wildlife Sanctuary, Eastern India using light use efficiency (LUE) models namely vegetation photosynthesis model (VPM) and vegetation photosynthesis and respiration model (VPRM). Several parameters like photosynthetically active radiation (PAR), fraction of absorbed photosynthetically active radiation (FAPAR), environmental scalars, etc. were estimated using the Sentinel 2 and Sentinel 3 satellite datasets to estimate primary productivity of the wildlife sanctuary. The study showed that the higher GPP in the month of June was found, primarily in the core zone of the sanctuary (29.1 to 15.3 g C m −2 d −1 ), which is dominated by species viz., Terminalia tomentosa , Syzygium cumini , Schleichera oleosa , Holarrhena pubescens , Mallotus philippensis , Terminalia bellirica , Cleistanthus collinus , and Holarrhena floribunda. While the lower GPP was observed in the month of February, in the parts of the core zone of the sanctuary (7.6 to 4.4 g C m −2 d −1 ). A higher GPP (20.3 g C m −2 d −1 ) was recorded in the pure patches of Shorea robusta during the month of June. Spatially, the core zone of the sanctuary was highly productive, due to dense forest cover and insignificant anthropogenic influence. The outer margins, especially the western parts of the Dalma wildlife sanctuary, experienced significant degradation because of fuelwood collection, which led to reduced primary productivity. The study demonstrated that productivity had a strong positive correlation with FAPAR ( R 2 = 0.91), and with field-measured leaf area index (LAI) ( R 2 = 0.79), while it has moderate positive correlation with (Enhanced vegetation index) EVI ( R 2 = 0.40), and PAR ( R 2 = 0.44) in the Tropical Dry Deciduous Forest of Eastern India. LICOR sensor-based field measured LUE showed high variation from one species to another in the field, suggesting critical assessment of dominant species. High spatial resolution of Sentinel-2 data allowed depiction of the GPP variance effectively in heterogeneous areas along with vital attributes of various environmental variables at finer scales (spatial and temporal) that support temporal assessment of carbon dynamics in the study area for effective forest conservation and management.