Climatic Drivers of Kharif Rice Yield Variability in Namsai, Arunachal Pradesh, India: Assessing the Role of Seasonal Rainfall and Temperature (2018–2023)

Abstract

Background: Climate variability is already affecting agricultural productivity, and the probability of severe yield losses is expected to increase under future climate scenarios Aim: This study quantifies climatic drivers of Kharif-season rice yield variability in Namsai district, Arunachal Pradesh. Study Design: The present study conducted using seasonally aggregated rainfall and temperature indicators with regression-based attribution and scenario modelling.    Place and Duration of Study: This study focused on the Namsai district of Arunachal Pradesh in Northeast India, where rice constitutes the dominant staple crop and agricultural production is largely dependent on monsoon rainfall. The present study was conducted during 2018–2023 using seasonal climate data obtained from the Indian Meteorological Department (IMD) Data Service Portal. Methodology: The climate data were aggregated across the Kharif growing season (June–October) to capture hydroclimatic conditions relevant to key rice growth stages. In the present study, bivariate association of rice yield with rainfall and with temperature was conducted, while also estimating conditional (partial) associations that net out the covariance between rainfall and temperature. Further, structural multivariate specification used to generate mean yield projections under counterfactual climate perturbations relative to a recent baseline period. Results: Seasonal rainfall varied markedly (2455.6–4545.8 mm), while mean temperature remained relatively stable (26.8–27.8°C). Rice yield exhibited substantial fluctuations (3.19–3.44 t ha⁻¹) that closely aligned with rainfall anomalies. Rainfall showed a moderate positive correlation with yield (r = 0.68), while temperature displayed a moderate negative association (r = −0.38). Regression results consistently identified rainfall as the dominant climate driver, with higher seasonal rainfall significantly reducing yield across bivariate, conditional, and multivariate models, whereas temperature effects were not statistically robust after controlling for rainfall. The joint climate model explained 51% of inter-annual yield variation (R² = 0.518). Scenario projections anchored to recent climate conditions indicated a baseline yield of about 3.36 t ha⁻¹, with lower expected yields under drier and warmer conditions. Conclusion: From a policy perspective, the results emphasize the need to prioritize rainfall-related adaptation strategies, including improved irrigation, drought-resilient varieties, flexible cropping calendars, and enhanced early-warning systems. Overall, it concludes that rainfall variability currently governs climate sensitivity in Namsai’s rainfed rice systems more strongly than temperature, offering valuable insights for designing targeted climate-resilient agricultural strategies in monsoon-driven regions.