Identifying the influence of heavy precipitation and ecological water replenishment (EWR) on groundwater resources is essential for the management of groundwater resources and for risk prevention. This study innovatively developed a groundwater resource analysis and prediction model integrated with the water level fluctuation method, correlation analysis, and machine learning method under the influence of heavy precipitation and EWR. The results of the water level fluctuation method showed that compared with January 1, 2021, the groundwater resources of the study area increased to 4.46 × 108 m3 on August 28. Compared with the small flow of EWR, heavy precipitation was the main contributor to the rise in the groundwater level. Correlation analysis found that elevation, specific yield, and permeability coefficient show positive correlations with groundwater resource recharge. Machine learning results showed that among the water level prediction models of 35 monitoring wells, extreme gradient boosting (XGB) and random forest (RF) performed best in 30 wells and 5 wells, respectively. The increase in groundwater storage predicted deviation from the actual value by only 0.6 × 107 m3 (prediction bias of 1.3%), indicating that the model prediction performance was well under the condition of heavy precipitation. This study can help to better understand the changing trends of groundwater resources under the influence of heavy precipitation and EWR.

  • Through hydrology, statistics, and machine learning, the groundwater changes under the dual effects of heavy precipitation and ecological water replenishment are studied.

  • A machine learning model is developed to predict the groundwater level and storage under heavy precipitation scenarios.

  • XGB and RF models well predicted the groundwater change the next day, and the prediction deviation was only 1.3%.

This content is only available as a PDF.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY-NC-ND 4.0), which permits copying and redistribution for non-commercial purposes with no derivatives, provided the original work is properly cited (

Supplementary data