Buffer effect on identifying transient streambed hydraulic conductivity with inversion of flood wave responses

Abstract

Determination of streambed hydraulic conductivity (Ksb) and its variations over time are critical for quantifying exchange fluxes between surface water and groundwater. Previous studies have used flood wave response techniques (FRT) to determine Ksb or its transience, excluding the streambed storage effect, thus the estimated Ksb may not be reliable. In this study, after improving a previous FRT by substituting the gradient descent algorithm with an advanced and robust optimization method, trust region reflective algorithm, we have analyzed the influence of streambed storage effect on the estimation of time-dependent Ksb through successive inversion of flood wave responses. The study shows that the streambed storage effect is not important for identifying the transience in the hydraulic properties of a relatively thin streambed (e.g., several or dozens of centimeters). But significant under-estimation (1–2 orders of magnitude) can be seen because of neglecting the streambed storage effect. It is particularly true for a stream-confined aquifer system if the streambed is composed of fine deposits (e.g., silt, clay) with a thicker thickness greater than 1 m corresponding to low permeability but high storativity, and the aquifer is characterized by relatively low permeability (e.g., roughly 0.1–10 m d−1) and relatively low storativity (roughly 1E-5 m−1) (e.g., dense fine sands). The new method in which streambed storage effect is incorporated, can provide more reliable estimations of time-dependent Ksb for a stream-confined aquifer system, and it has been successfully applied to estimate the time-dependent Ksb of a reach of the Arkansas River, Kansas, USA.