Author(s): Byunguk Kim; Il Won Seo

Linked Author(s): Byunguk Kim, Il Won Seo

Keywords: Solute transport; Non-Fickian transport; Net retention time; Tracer test; Transient storage model

Abstract: Solute transport analysis in rivers is the fundamental work for swift countermeasure against pollutant spill accidents into the river system. This transport is dominated not only by hydrodynamics within surface flow domain, but also by interactions with surrounded storage zones in rivers. In this respect, the Transient Storage Model (TSM) has been most widely used due to its simple conceptualization in which the storage mechanism is characterized as equivalent to a first-order mass transfer. However, this simplicity has been commented as a deficiency due to poor approximation in tailing behavior of solute cloud and corresponding parameter uncertainty. In the sense that such tailing behavior in a breakthrough curve is most representable part for storage mechanism, the validity of most storage system modeling were demonstrated by comparing the tails. However, storage effect is, in fact, less responsible in generating breakthrough curves than surface flow dynamics. To identify the storage system, investigation of actual solute behavior in storage zones is required. Unfortunately, it is difficult in streams due to the diversity, invisibility, and unclear boundary of natural storage zones This study aimed to mathematically estimation how long solute particles were trapped within storage zones along a stream reach using deconvolution technique with observed breakthrough curves. To obtain the observation in a natural stream, the tracer test was carried out at the 4.54 km reach of Gam Creek, consisting primarily of sand substrate, South Korea in 2020. The problematic high frequency variability was removed by the Butterworth filter. The undesired oscillating problem was controlled by configuring the filtering order parameter. As a result, we successfully estimated the probability density function of net retention time of solute without significant distortion in primary breakthrough curves. Using the net retention time distribution, we evaluate the performance of the TSM simulation. One of the key results is that the simulation results of the parameter-optimized TSM underestimated the storage effect by as much as an average 44 %, despite the high accuracy with 0.99 of R² in breakthrough curve approximation.

DOI: https://doi.org/10.3850/IAHR-39WC252171192022406

Year: 2022