Abstract
Evapotranspiration is a key component of water and energy balance. An accurate estimate of reference evapotranspiration (ETo) is important for determining the water demand of field crops, water management, and hydrological modelling. The FAO-PM ETo equation is the standard equation for estimating ETo. Its use is limited by the requirement for too many observed inputs that are not always available in most weather stations. Empirical models requiring readily available inputs have been developed as an alternative. However, their performance is location-specific. Therefore, this study assesses the performance of the FAO-PM ETo computed with limited data and fourteen empirical ETo models. Meteorological data (2012–2019) was analyzed under the semi-arid climate conditions in southern Manitoba. Model performance was assessed using statistical indices, including R2, RMSE, NSE, MPE, and MAE. Results showed that ETo estimates under missing wind speed, relative humidity, or solar radiation were acceptable, although model performance decreased with increased missing data yielding average to poor performance. Based on R2, RMSE, and NSE values, among the 14 models compared, the best performing models are Valiantzas-1, Valiantzas-3, Irmak, Valiantzas-2, and Priestly-Taylor models. New empirical coefficients were developed, requiring one or two climatic inputs to improve empirical models. Results showed that the calibrated models performed better than the original equation. The sensitivity analysis showed that ETo is most sensitive to maximum temperature (Tmax) and solar radiation (Rs), followed by vapour pressure deficit (VPD), wind speed (U2), and minimum temperature (Tmin). Therefore, it is recommended to ensure accurate measurements of temperature and solar radiation for accurate ETo estimates. This study provides alternative ETo models with reasonable and accurate ETo estimates for southern Manitoba and other areas with similar climate characteristics.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have