|
|
Ensemble learning estimation of reference crop evapotranspiration taking into account spatial and temporal characteristics |
LIU Ao1, ZHAO Dongbao1*, WEI Yichang1, XIAO Lian2 |
1. College of Surveying and Geo-Informatics, North China University of Water Resources and Electric Power, Zhengzhou, Henan 450046, China; 2. Sichuan Basic Geographic Information Center, Ministry of Natural Resources, Chengdu, Sichuan 610041, China |
|
|
Guide |
|
Abstract In order to improve the estimation accuracy of reference crop evapotranspiration(ET0), this study takes Sichuan Province as the research area, and reveals that the data changes of ET0 in Sichuan Pro-vince have obvious temporal and spatial autocorrelation. Then, based on meteorological characte-ristics, spatio and temporal features are introduced to construct XGBoost and lightGBM, GBDT, random forest and extreme tree-based Stacking model. The Stacking model with its base model and empirical model(FAO 56 Penman-Monteith)was comprehensively verified from multiple indicators such as coefficient of determination, mean absolute error, and mean square error. The experimental results show that considering the spatial characteristics, the Stacking model improves the coefficient of determination on the test set by 3%, and the mean absolute error and mean square error are reduced by 51% and 76%, respectively. Taking into account the timing characteristics, the Stacking model improves the accuracy of the upper coefficient of determination on the test set by 4%, and the mean absolute error and mean square error are reduced by 92% and 72%, respectively. This shows that the introduction of spatiotemporal features can effectively improve the ET0 performance of the model. With both spatial and timing characteristics in mind, compared with Penman-Monteith′s formula, the Stacking model′s coefficient of determination increased by 39%, and the mean absolute error and mean square error reduced by 95% and 77%, respectively, and in the year-by-year accuracy verification in 2006—2010, the accuracy of the Stacking model is always better than its annual accuracy optimal base model. Therefore, the Stacking model considering the spatial and timing characteristics in this paper can effectively improve the ET0 estimation accuracy in Sichuan Province.
|
Received: 06 April 2023
|
Fund: |
|
|
|
|
|
|