Among a number of methodologies for developing groundwater supply to overcome drought events reported in the research community, an accurate estimation of the groundwater level is an important initial issue to provide an efficient method for operating groundwater. The primary objective of this paper is to develop an advanced prediction model for the groundwater level in the catchment area of the Ssangcheon groundwater dam using precipitation based period dividing algorithm and response surface methodology (RSM). A numerical example clearly shows that the proposed method can effectively forecast groundwater level in terms of correlation coefficient ($R^2$) in the upstream part of the Ssangcheon groundwater dam.