Characterizing coefficient of permeability based on response of groundwater level to river stage using regional database

The coefficient of permeability ( k ) is an important soil property for various flow-related problems and site characterization. In this study, a new method for estimating k is proposed based on the time response of groundwater level (GWL) to river stage (RS) with changes in loosing and gaining streams during normal and rainy periods, respectively. Focus was given on establishing a method that does not require costly experimental testing procedure while utilizing locally available hydrology and geologic database. For this purpose, the GWL–RS correlation model was introduced and adopted to develop the k correlation model based on the time response of GWL to RS. The proposed method was calibrated using measured datasets of GWL and RS collected from 17 different regions. Based on the compared results from the measured datasets, a modification factor ( M f ) was established and adopted to further enhance the k correlation model. When the M f was adopted, the R 2 value of the prediction was 0.8756, and the RMSE decreased to 0.00643. To check the applicability of the proposed method, additional datasets were collected and adopted into the comparison. It was confirmed that measured and predicted soil permeabilities were in close agreement.