Satellite-based estimation of potential evapotranspiration using the Thornthwaite–Mather model for sub-regional water resource assessment
摘要
Accurate estimation of potential evapotranspiration (PET) is crucial for agricultural planning and irrigation management, particularly for determining crop water requirements and irrigation scheduling. However, the unavailability of ground-based meteorological data poses a significant challenge, especially in Bihar, India. This study evaluates the spatio-temporal variability of PET over Samastipur district, Bihar, using the Thornthwaite–Mather model. Land surface temperature (LST) was derived from the GLDAS-2.1 Noah Land Surface Model L4 satellite product at a spatial resolution of 0.25° × 0.25° on a monthly timescale for 21 years (2000–2020). The satellite-derived LST was compared and validated with ground-based temperature data from the Meteorological Station (MS), Pusa, using statistical metrics. BIAS-adjusted LST demonstrated improved agreement with observed temperature, enhancing the R2 value from 0.89 to 0.96. Estimated PET values at 67 grid points showed strong correlation (R2 = 0.90) with reference evapotranspiration (ETo) recorded at MS, Pusa. The seasonal pattern exhibited maximum PET (120.7 mm) in June and minimum (5.5 mm) in January. Trend analysis using the Mann–Kendall test and Sen’s slope estimator revealed no significant monthly, seasonal, or annual changes in PET over 21 years at MS Pusa, indicating overall stability despite minor non-significant fluctuations.” The outcomes of this study offer valuable inputs for agricultural decision-making by delineating critical periods of crop water demand, facilitating the optimization of irrigation scheduling and crop calendar, and enhancing water-use efficiency. Consequently, the findings contribute to the development of resilient and sustainable agricultural planning frameworks, particularly in data scare regions.