ABSTRACT
The terrestrial biosphere plays a significant role in governing the climate system through regulating carbon fluxes between the vegetation and the atmosphere. Gross primary production (GPP) plays a critical role in understanding the carbon cycle and the biogeochemical process of terrestrial ecosystems. Numerous remote sensing models have been proposed to estimate GPP, which can be categorized into statistical models, light use efficiency models, and process-based models that heavily rely on environmental parameters. This study has employed a vegetation indices (VIs)–based model (i.e., VI × VI × PAR) to estimate daily GPP over cropland on Majuli Island using high spatial resolution PlanetScope satellite data (3.125m) and photosynthetically active radiation (PAR). Almost 15 VI-based models were developed, wherein the NDVI × CIgreen × PAR model was found to be better than others to deduce GPP accurately. The results showed that the daily GPP varied from 5 to 5.9 gC m−2 day−1 with a root mean square error (RMSE) of 0.67 gC m−2 d−1 in November (post-monsoon season). The NDVI × CIgreen × PAR model was comparable to MODIS GPP estimates, indicating a high correlation (R2 = 0.88, p-value < 0.001). The present study uses a very high-resolution LULC map and estimates the GPP of croplands, which may be useful to policymakers and agriculture managers to maximize the carbon assimilation and attain optimum crop yields.
