Dynamic groundwater level estimation by the velocity spectrum analysis of GPR

Abstract

Time-lapse Common Mid-Point (CMP) measurement using Ground Penetrating Radar (GPR) was carried out to investigate its ability of detecting and estimating the dynamic groundwater change. The survey was conducted repeatedly at a production well in a water source area near Tuul river in Ulaanbaatar, Mongolia, when the groundwater table was under pumping condition and static condition. A new algorithm which involves envelope velocity spectrum and an automatic velocity picking scheme was proposed. A vertical interval velocity profile can be accurately derived from a CMP dataset using the proposed algorithm. The velocity of the surface soil is estimated to be 0.161 m/ns, which agrees well with the estimation of 0.159 m/ns by Time Domain Reflectometry (TDR). The vertical velocity profile was converted to the vertical water content profile, in which the groundwater level could be easily determined from a feature of evident increase of water content. In this measurement, we observed a total groundwater level rise of 0.21 m under the mid-point of CMP survey line after the pumping was stopped in a non-destructive way. The dynamic groundwater level change estimated by GPR combining other hydrogeological information was input to an unconfined aquifer model to estimate the hydraulic conductivity of the unconfined groundwater aquifer. The result agreed well with a previous study. We think the hydraulic parameter estimation by GPR is much easier than the conventional method, in which we have to drill many boreholes.