Parameter regionalization based on machine learning optimizes the estimation of reference evapotranspiration in data deficient area
Zhangkang Shu,You Zhou,Jianyun Zhang,Junliang Jin,Lin Wang,Ningbo Cui,Guoqing Wang,Jiangjiang Zhang,Houfa Wu,Zongjun Wu,Xi Chen
DOI: https://doi.org/10.1016/j.scitotenv.2022.157034
2022-10-20
Abstract:Reference evapotranspiration (ET0), as one important variable in climatology, hydrology, and agricultural science, plays an important role in the terrestrial hydrological cycle and agricultural irrigation. However, the ET0 estimation process is inaccurate due to the lack of weather stations and historical data. In this study, a new method of ET0 estimation was proposed to improve the ET0 estimation performance in regions with limited data. Four empirical models with different data requirements, Albrecht, Hargreaves-Samani, Priestley-Taylor, and Penman, were applied and optimized the parameters by the Shuffled Complex Evolution-University of Arizona algorithm with the ET0 calculated by the Penman-Monteith model as the reference value at 600 meteorological stations in China. Two machine learning models, Random Forest (RF) and Multiple Linear Regression (MLR) were used to establish the regionalization of the parameter of the empirical model. The result showed that parameter optimization could significantly improve ET0 estimation in different climate regions of China. The Penman model has the strongest physical foundation and the highest estimation accuracy, followed by the Hargeaves-Samani and Priestley-Taylor model. The mass-transfer-based model, Albrecht, could only estimate regional ET0 efficiently after parameter optimization. Based on the more advanced RF machine learning regionalization method that considers complex linear relationships of variables, ET0 estimation in regions lacking data could be improved efficiently. Machine learning could be used to describe the ET0 model parameters in different regions because of the similarity. The combination of machine learning and empirical model could provide a new method for ET0 estimation in data deficient regions.
environmental sciences