Forecasting Multi-Step Ahead Monthly Reference Evapotranspiration Using Hybrid Extreme Gradient Boosting with Grey Wolf Optimization Algorithm

It is important for regional water resources management to know the agricultural water consumption information several months in advance.Forecasting reference evapotranspiration(ET_(0))in the next few months is important for irrigation and reservoir management.Studies on forecasting of multiple-month ahead ET_(0) using machine learning models have not been reported yet.Besides,machine learning models such as the XGBoost model has multiple parameters that need to be tuned,and traditional methods can get stuck in a regional optimal solution and fail to obtain a global optimal solution.This study investigated the performance of the hybrid extreme gradient boosting(XGBoost)model coupled with the Grey Wolf Optimizer(GWO)algorithm for forecasting multi-step ahead ET_(0)(1-3 months ahead),compared with three conventional machine learning models,i.e.,standalone XGBoost,multi-layer perceptron(MLP)and M5 model tree(M5)models in the subtropical zone of China.The results showed that theGWO-XGB model generally performed better than the other three machine learning models in forecasting 1-3 months ahead ET_(0),followed by the XGB,M5 and MLP models with very small differences among the three models.The GWO-XGB model performed best in autumn,while the MLP model performed slightly better than the other three models in summer.It is thus suggested to apply the MLP model for ET_(0) forecasting in summer but use the GWO-XGB model in other seasons.

Explaining the evaporation paradox in Jiangxi Province of China:Spatial distribution and temporal trends in potential evapotranspiration of Jiangxi Province from 1961 to 2013

Evaporation acts as an important component and a key control factor in land hydrological processes.In order to analyze the trend of change on potential evapotranspiration from 1961 to 2013 and to discuss the existence of the evaporation paradox in Jiangxi province,China,monthly meteorological data spanning the years 1961–2013 were analyzed in this study,where the data were collected from 15 national meteorological stations in Jiangxi Province.The Penman–Monteith equation was employed to compute the potential evapotranspiration(ET0).Spatial interpolation and data mining technology were used to analyze the spatial and temporal changes of ET0 and air temperature,with the effort to explain the evaporation paradox.By solving the total differential and the partial derivatives coefficients of the independent variables in Penman–Monteith equation,the cause of the paradox was quantitatively evaluated.The results showed that the annual ET0 had been decreasing significantly in Jiangxi Province since 1979,whereas the air temperature had been rising significantly,presenting the evaporation paradox.The decreases in sunshine duration and wind speed reduced ET0 by 0.207 mm and 0.060 mm,respectively,accounting for 92.3%and 26.7%of the total ET0,respectively.It is concluded that sunshine duration and wind speed are the main causes to the decrease in potential evapotranspiration in Jiangxi Province.