Correlation-based analysis of weighted and unweighted weather indices for rice yield in Kashmir

Abstract

Rice is a vital staple crop in the Kashmir division, where climatic variability poses significant challenges to agricultural productivity. This study investigates the relationship between rice yield and weekly weather variables: maximum and minimum temperature, rainfall, sunshine hours, and relative humidity over a 39-year period (1985-2023). Using meteorological data from the Indian Meteorological Department (IMD) and yield records from the Directorate of Economics and Statistics, a comprehensive correlation analysis using Pearson’s correlation method was conducted based on both unweighted and weighted weather indices.The findings reveal that weighted weather indices exhibit significantly stronger correlations with rice yield compared to their unweighted counterparts, underscoring their superior predictive value. Among individual variables, rainfall (Z31) and relative humidity (Z51) emerged as the most influential, while interaction-based weighted indices such as Z131 demonstrated exceptionally high correlations (r > 0.70). Visual tools like correlation heatmaps and bar plots further validated the strength and direction of these associations.