Electronic Theses and Dissertations

Date of Award


Document Type


Degree Name

Ph.D. in Engineering Science

First Advisor

Mohammad Al-Hamdan

Second Advisor

Yavuz Ozeren

Relational Format



Streambed conductivity is widely recognized as a vital parameter controlling the stream-aquifer exchange. Many field measurements have revealed that temporal variation of streambed conductivity is rapid and dramatic during floods. However, previous numerical studies either assumed streambed conductivity as steady over storm events or focused on time-varying streambed conductivity during low-stage periods. This study aims at bridging this gap by both incorporating time-varying streambed conductivity into a numerical model of groundwater (GW) – surface water (SW) interactions and simulating them under different flooding conditions, making this dissertation the first modeling study to explore combined effects of high-stage events and time-varying streambed conductivity on GW-SW exchanges.

To this end, a finite element numerical model, CCHE3D-GW, which can simulate both saturated and unsaturated GW flows, was developed, and thoroughly verified in this dissertation in order to facilitate numerical studies. CCHE3D-GW was then used for the calibration of a pumping test conducted in the study area, Money experimental site in Mississippi, USA, in order to obtain the hydrogeological information.

A rapid model that can estimate time-varying streambed conductivity based on GW responses to flood-wave fluctuations was developed, and then validated with two typical synthetic cases. Its applicability to a heterogeneous aquifer was tested and confirmed as well. The sensitivity analysis showed that the rapid model was approximately linearly affected by the uncertainty of its input parameters. The rapid model was then applied to obtain the time series of the streambed conductivity for the Money experimental site, which is near the Tallahatchie River. Three typical flooding conditions were analyzed: 1) one that consisted of multiple flood events with attenuating amplitudes, 2) one that was composed of multiple flood events with amplifying amplitudes and 3) one with a single flood event (the river stage reached a historically high value).

The estimated time-varying streambed conductivity was then imposed on CCHE3D-GW to study two commonly encountered cases of GW-SW interactions. The first case was a hypothetical riverside pump in the Money experimental site, and the second case was the dune-induced hyporheic exchange. The simulations were conducted for the aforementioned three flooding conditions with both the static (traditional manner) and time-varying streambed conductivity.

Based on the modeling results and analyses for different flooding conditions in this dissertation, it can be concluded that considering time-varying streambed conductivity is imperative for accurately understanding the evolution of the GW-SW exchange intensity and residence time of the infiltrated water. Hence, demonstrating that time-varying streambed conductivity is a critical and essential variable for accurate and robust process-based GW-SW interaction modeling, and particularly during flooding periods.


Hydraulic engineering

Available for download on Friday, February 17, 2023