Reference Evapotranspiration Modeling for Northwestern Bangladesh Using Multiple Linear Regression

Abstract

Abstract The estimation of reference evapotranspiration (ETo) is crucial for determining crop evapotranspiration rates. The FAO-56 Penman-Monteith (FAO-56 PM) method is globally recognized for its consistent performance in ETo estimation. However, alternative methods remain popular due to their simplicity and minimal data requirements. This study investigated ETo modeling using multiple linear regression for three stations in northwestern Bangladesh: Rajshahi, Bogra, and Rangpur. Analyzing meteorological data from 1979 to 2022, the study explored relationships between ETo and various meteorological variables. Solar radiation consistently showed the strongest correlation (τ > 0.7) with ETo, followed by maximum temperature (τ > 0.65), while sunshine hour and relative humidity displayed weaker correlations (τ < 0.1) at every station. Three regression models were developed, all with r2 values > 0.93 and RMSE < 0.07. Model 3, including average temperature and relative humidity alongside solar radiation and wind speed, outperformed others (r2 > 0.97 and RMSE < 0.025). The findings have significant implications for agricultural planning, water resource management, and climate change adaptation in the region. Accurate ETo predictions enable stakeholders to optimize irrigation scheduling, improve crop yield forecasts, and formulate resilient agricultural policies.