Forecasting daily reference evapotranspiration for Canada using the Penman–Monteith model and statistically downscaled global climate model projections
Xiaohui Yan, Abdolmajid Mohammadian
Abstract
The forecasting performance for daily reference evapotranspiration (ET0) at six weather stations in Canada using the Penman–Monteith (P–M) model and statistically downscaled global climate model (GCM) projections was evaluated and quantified. The observational daily weather data for ET0 calculations, including the daily air temperature (Ta), solar radiation (Rs), wind speed (Uw), and relative humidity (RH), were collected. The GCM data obtained by the ESM2M model for the same variables, period, and stations were extracted and downscaled using the Quantiles–matching approach. The forecasting performance for the weather variables was quantified, and the results showed that the raw GCM data for Rs and Ta over Canada matched the observations very well, but the GCM performed relatively worse in forecasting Uw and RH. The Quantiles–matching downscaling approach can significantly improve the forecast accuracy of the meteorological variables. The ET0 calculated using the raw GCM or downscaled GCM data were then compared with those computed with the observational data. The results demonstrated that the daily ET0 over Canada can be satisfactorily forecasted using the P–M model and statistically downscaled GCM projections.