ESTIMATING HYDRAULIC PROPERTIES OF A RIVER BANK AQUIFER UNDER TIDAL INFLUENCE

Quantifying water exchange between rivers and aquifers is crucial to understand the fate of heavy metals and thus demands accurate estimates of aquifer properties. Intensive irrigation pumping in riverbank aquifers during dry seasons induces river water intrusion into the aquifer and further aggregates solute migration. This paper presents and compares multiple independent methods to estimate the hydraulic properties of a shallow riverbank aquifer bounded by a large tidally influenced river in Bangladesh. Slug tests gave estimates of aquifer hydraulic conductivity (K) ranging from 18 to 34 m/d. The aquifer dimensions were confirmed using borehole lithologies and an electrical resistivity tomography (ERT) survey. Fast Fourier transform (FFT) was used to identify the major tidal modes within the river and well hydrographs from which amplitude attenuations and time lags were calculated. Aquifer diffusivity (D) was estimated using the Jacob-Ferris propagation model assuming one dimensional confined aquifer. The diffusivity values estimated from amplitude attenuation were consistently larger than those estimated from time lag method. Time lags and amplitude attenuations observed in the monitoring wells were used to correct drawdown curves from a tidally influenced pumping test. The optimal transmissivity (T) and storativity (S) values were obtained through 1D inverse modeling using the tidal-corrected drawdown curve data. This yielded values of T and S ranging from 400 to 500 m²/d and 0.0001 to 0.0005 [-], respectively. Estimated diffusivity from tidal methods was approximately an order of magnitude lower compared to the pumping test results. To resolve this discrepancy, 2D numerical simulations were conducted to test the effect of bank morphology, aquifer anisotropy and heterogeneity on sinusoidal wave propagation into river bank aquifers. It was concluded that bank slope angle and aquifer heterogeneity yielded similar effects on wave propagations. However, aquifer anisotropy produced the least damped and lagged wave signal but caused the largest propagation bias. It is expected this analysis will lead to a set of optimal semi-analytical heuristic equations for estimating aquifer properties in different physical settings.