India constitutes one of the preeminent global producers of rice, representing approximately 20% of the total rice production worldwide. The cultivation of rice occupies nearly 27.6% of country arable land, with an average per capita consumption of roughly 68.2 kg of milled rice per annum. As a fundamental staple food, it is of paramount significance for policymakers, planners, and researchers to possess a precise estimate prior to the harvest of crops. The timely acquisition of accurate statistics facilitates planners and decisionmakers in devising policies concerning import and export strategies in instances of deficit or surplus. The present investigation seeks to assess the viability of remote sensing methodologies for estimating yields in the primary rice-producing districts of eastern Vidarbha region of Maharashtra. Recent advancements in remote sensing resolutions-encompassing spectral, spatial, radiometric, and temporal aspects-have enabled the timely collection of pertinent information. This research developed an intermediary approach termed the semi-physical method, which integrates remote sensing data with physiological concepts such as Photosynthetically Active Radiation and the fraction of PAR assimilated by the crop. The computation of Net Primary Productivity (NPP) was executed utilizing the Monteith model. The estimation of rice yield was derived from the actual NPP, radiation use efficiency, and harvest index. The study was conducted during the kharif rice-growing season (June-November) for the year 2022. Although the model presents a slight discrepancy in yield when compared to both the actual (government-observed) and modelled (estimated) figures, the results indicate that the estimated yield is +9.5% higher than the district-level actual observed value obtained from the Maharashtra Department of Agriculture (https://krishi.maharashtra.gov.in). The research illustrates that this methodology can yield estimations at sub-district levels or smaller administrative units; however, attaining precision at a more granular scale necessitates supplementary site-specific data.
Meshram et al. (2025) studied this question.