AUTHOR=Guo Benli , Zhang Shouchuan , Liu Kai , Yang Peng , Xing Honglian , Feng Qiyuan , Zhu Wei , Zhang Yaoyao , Jia Wuhui 

TITLE=Prediction of groundwater level under the influence of groundwater exploitation using a data-driven method with the combination of time series analysis and long short-term memory: a case study of a coastal aquifer in Rizhao City, Northern China

JOURNAL=Frontiers in Environmental Science

VOLUME=Volume 11 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/environmental-science/articles/10.3389/fenvs.2023.1253949

DOI=10.3389/fenvs.2023.1253949

ISSN=2296-665X

ABSTRACT=The excessive exploitation of groundwater not only destroys the dynamic balance between coastal aquifer and seawater, but also causes a series of geological and environmental problems. The groundwater level prediction provides an efficiency way to solve these intractable ecological problems. Due to the characteristic of spatially and temporally complex hydrologic processes, although several hydrological numerical models have been employed to conduct prediction, no study has accurately predicted the groundwater level change under the consideration of groundwater exploitation, especially in the coastal aquifer. This study proposes the novel data-driven method based on the combination of times series analysis and the machine learning method for predicting the variation of groundwater level in the coastal aquifer under the influence of groundwater exploitation accurately. The partial autocorrelation function and continuous wavelet coherence are first to analyze the monitoring data of groundwater level at three wells, which indicate that the historical monitored data and the dataset of precipitation can be considered as the input variables to construct the hydrological model. Then, three models based on the different inputs are constructed, including LSTM, PACF-LSTM, and PACF-WC-LSTM. The performances of three models are compared by the calculation of four error metrics. The results show that the performance of PACF-LSTM and PACF-WC-LSTM are better than LSTM, and the PACF-WC-LSTM model achieved the best prediction performance.Accurately predicting the variation of groundwater level provide the basis for managing groundwater resources and preserving ecological environment.