Prediction of groundwater levels by 2100 considering climate change based on precipitation and river stage

Abstract

Abstract Changes in GWL is an important consideration for establishing sustainable social infra system and disaster prevention. In this study, methodology for predicting future GWL is established based on the climate change scenario and adopted to predict GWL by year 2100. Deep neural network (DNN) and engineering groundwater level prediction model (EGPM) were adopted for precipitation- and river stage-dominant cases. Database of hydrological and geographical parameters was established and introduced into the prediction. It was shown that present and future GWLs were not the same, and the difference was most noticeable in near-river area. The increases in GWL by 2100 in Magok region was estimated around 1.8 m. The GWL in Jangwi and Hangdong regions indicated an increase around 0.18 m, implying that the increase in GWL would not be dramatically large, yet gradual increase is anticipated. All these indicated a tendency of higher GWL in future with magnitude depending on regional and locational conditions of target area. Predicted future GWL profiles were introduced into the 3D flow analysis program MODFLOW to establish a visual utilization of predicted GWL.