Understanding groundwater table using a statistical model

Abstract In this study, a statistical model was established to estimate the groundwater table using precipitation, evaporation, the river stage of the Liangduo River, and the tide level of the Yellow Sea, as well as to predict the groundwater table with easily measurable climate data in a coastal plain in eastern China. To achieve these objectives, groundwater table data from twelve wells in a farmland covering an area of 50 m × 150 m were measured over a 12-month period in 2013 in Dongtai City, Jiangsu Province. Trend analysis and correlation analysis were conducted to study the patterns of changes in the groundwater table. In addition, a linear regression model was established and regression analysis was conducted to understand the relationships between precipitation, evaporation, river stage, tide level, and groundwater table. The results are as follows: (1) The groundwater table was strongly affected by climate factors (e.g., precipitation and evaporation), and river stage was also a significant factor affecting the groundwater table in the study area ( p p is the probability value). (2) The groundwater table was especially sensitive to precipitation. The significance of the factors of the groundwater table were ranked in the following descending order: precipitation, evaporation, and river stage. (3) A triple linear regression model of the groundwater table, precipitation, evaporation, and river stage was established. The linear relationship between the groundwater table and the main factors was satisfied by the actual values versus the simulated values of the groundwater table ( R 2  = 0.841, where R 2 is the coefficient of determination).