Predicting Groundwater Level Based on Machine Learning: A Case Study of the Hebei Plain

In recent years, the groundwater level (GWL) and its dynamic changes in Hebei Plain have gained increasing interest. The GWL serves as a crucial indicator of health resources, accurately predicting is vital to prevent overexploitation loss water quality land subsidence. Here, we utilized data-driven models, such support vector machine, long-short term memory, multi-layer perceptron, gated recurrent unit predict GWL. Additionally, data from six monitoring stations 2018 2020, covering dynamical fluctuations, increases, decreases GWL, were used. Further, first 70% remaining 30% time-series used train test model, respectively. Each model was quantitatively evaluated using root mean square error (RMSE), coefficient determination (R2), Nash–Sutcliffe efficiency (NSE), they qualitatively line plots, scatter Taylor diagrams. A comparison models revealed that RMSE, R2, NSE GRU training testing periods better than those other at most stations. conclusion, performed best could predictions Plain.