Groundwater level prediction using a SOM-aided stepwise cluster inference model

Accurate groundwater level (GWL) prediction can contribute to sustaining reliable water supply to domestic, agricultural and industrial uses as well as ecological services, especially in arid and semi-arid areas. In this paper, a regional GWL modeling framework was first presented through coupling both spatial and temporal clustering techniques. Specifically, the self-organizing map (SOM) was applied to identify spatially homogeneous clusters …