Hybrid deep learning method for a week-ahead evapotranspiration forecasting

• Published in: Stochastic environmental research and risk assessment Vol. 36; no. 3; pp. 831 - 849
• Main Authors: Ahmed, A. A. Masrur, Deo, Ravinesh C., Feng, Qi, Ghahramani, Afshin, Raj, Nawin, Yin, Zhenliang, Yang, Linshan
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.03.2022; Springer Nature B.V
• Subjects: Ant colony optimization; Aquatic Pollution; Artificial neural networks; Atmospheric models; Atmospheric water; Balance studies; Chemistry and Earth Sciences; Computational Intelligence; Computer Science; Datasets; Deep learning; Drought; Earth and Environmental Science; Earth Sciences; Environment; Evapotranspiration; Forecasting; Hydrology; Math. Appl. in Environmental Science; Mathematical models; Neural networks; Original Paper; Physics; Probability Theory and Stochastic Processes; Soil water; Spectroradiometers; Statistics for Engineering; Waste Water Technology; Water balance; Water Management; Water Pollution Control; Gated recurrent unit; Hybrid-deep learning; Convolutional neural network; forecasting; ET
• ISSN: 1436-3240; 1436-3259
• DOI: 10.1007/s00477-021-02078-x

Reference crop evapotranspiration (ET o ) is an integral hydrological factor in soil–plant-atmospheric water balance studies and the management of drought events. This paper proposes a new hybrid-deep learning approach, combining convolutional neural network (CNN) and gated recurrent unit (GRU) along with Ant Colony Optimization (ACO), for a multi-step (week 1 to 4) daily-ET o forecast. The method also assimilates a comprehensive dataset with 52 diverse predictors, i.e., satellite-derived moderate resolution imaging spectroradiometer, ground-based datasets from scientific information for landowners and synoptic-scale climate indices. To develop a vigorous CNN-GRU model, a feature selection stage entails the ant colony optimization method implemented to improve the ET o forecast model for the three selected sites in Australian Murray Darling Basin. The results demonstrate excellent forecasting capability of the hybrid CNN-GRU model against the counterpart benchmark models, evidenced by a relatively small mean absolute error and high efficiency. Overall, this study shows that the proposed hybrid CNN-GRU model successfully apprehends the complex and non-linear relationships between predictor variables and the daily ET o .