University of Texas at Austin


Research Shows Proposed Changes to Texas Ship Channel Depth has Minimal Storm Surge Effect

By Joanne Foote

Published Oct. 13, 2023

The natural inlet and ship channel splits San José Island and Mustang Island, near the Port Aransas Ferry crossing. Credit: Adobe Stock image.

A new case study using computational modeling has predicted minimal effects on increasing hurricane storm surge in surrounding Texas Gulf Coast communities as a byproduct of proposed depth increase of the Corpus Christi Ship Channel near Aransas Pass.

The study, published in the Journal of Computational Science, focused on the portion of the ship channel near Aransas Pass, a primary access point for commercial vessels to enter and exit the system of bays between the Gulf of Mexico and the Port of Corpus Christi. This natural tidal inlet divides Mustang Island and San José Island, and are protective barrier islands to mainland Gulf Coast communities. Researchers at the Oden Institute for Computational Engineering and Science at The University of Texas at Austin utilized computing resources at the Texas Advanced Computer Center (TACC) for their findings.

The Port of Corpus Christi located south of Houston is the nation’s largest U.S. energy export gateway and third largest seaport in total waterway tonnage in the country, according to their website. As part of a larger project to accommodate industry growth, it has been proposed to deepen the ship channel through Aransas Pass near Harbor Island to 70 ft. (21.33 m), from its current average depth of 47 ft. (14.33 m), which would allow for very large crude carriers (VLCCs) to access the port more easily. However, there is community concern that a deeper channel could affect the seawater flow through the channel and the bays behind it, which has environmental, recreational, and commercial interests for the surrounding communities.

Research associate Eirik Valseth, and Clint Dawson, Director of the Computational Hydraulics Group (CHG) at the Oden Institute, and Department Chair of Aerospace Engineering and Engineering Mechanics at the Cockrell School, collaborated with Edward Buskey, Director of The University of Texas Marine Science Institute in Port Aransas, Texas to predict what effect deepening the channel could have on area storm surge during a hurricane. Valseth emphasized that this research is independent of the current ship channel expansion project.

Had the model results suggested a strong increase in storm surge, the proposed deepening of the ship channel could have led to very negative effects, including increased risk of flooding.

— Eirik Valseth, Oden Institute Research Associate

Their investigation is based on numerical mathematical models of the circulation of coastal water due to Hurricane Harvey which impacted the Texas coast in August 2017 and a synthetic hurricane derived from Harvey.

According to Valseth – whose research focuses on flood modeling, including storm surge – the computational model results were counter to what was initially suspected, which was that a deeper channel would allow more water to enter the region behind Aransas Pass. 

“In most of the region we noticed that a deeper channel would reduce the peak storm surge based on the model results. Closer investigation of hurricane wind fields and the results showed that indeed more water was allowed to pass through the channel, however, this is a two-way street. A significant amount of surge originates in the large shallow bays behind the Aransas Pass due to the counterclockwise rotating wind field of a hurricane. Hence, the deeper channel is also able to drain out this surge more easily than a shallow channel,” he said. 

This latest work piggybacks off the initial study which researched the effects of channel expansion on red drum larvae. According to Valseth, in the past work, a significant part was the development of the computerized descriptions of the physical domain, i.e., the finite element meshes used. 


“The meshes (or models) were identical except for the depth of the ship channel so we were able to predict the impacts of the deeper channel. We used these same models to perform the current study of hurricane storm surge impacts.” Numerical models developed and used for this and the previous projects requires high resolution finite element meshes of the region near the ship channel itself as well as the adjacent coastal ocean and the entire northern Atlantic Ocean. 

“The reason is that the impacts of a hurricane making landfall in Texas are greatly influenced by the flow of seawater during the entire hurricane process from build-up in the Atlantic, to making landfall in Texas. In fact, without this resolution and extent, the numerical predictions we made would have been of significantly reduced quality and accuracy. The models contain several millions of unknowns that must be computed and recomputed every second of simulated time,” Valseth stated.

“Had the model results suggested a strong increase in storm surge, the proposed deepening of the ship channel could have led to very negative effects, including increased risk of flooding,” said Valseth. However, the abstract from the study says that while the model results indicate that the changes to maximum storm surge magnitude are small and in large portions of the study area the models indicate a reduction in surge magnitude, there are some local areas where the models results show an increase of up to 30 centimeters in a synthetic extreme storm scenario.


Valseth stressed that these results are from computer modeling, noting that these will never reproduce the physics with 100% accuracy. However, as part of the development procedure the research team performed a validation of the model outputs by comparing to National Oceanic and Atmospheric Administration (NOAA) gauge data in the area studied. This validation produced similar accuracy as observed in past publications and the model was considered to be sufficiently accurate. 

“The key objective here was to compare model results for two distinct channel depths while all other inputs were kept unchanged. The underlying models are currently also used in the operational forecasting of hurricane storm surge on the Texas coast and are re-validated on a regular basis,” Valseth said.

“A significant number of model runs was necessary, and the rapid turnaround time of Frontera computations allowed us to continuously and effectively communicate results with the UTMSI collaborators,” said Valseth. To put that into perspective, approximately 3000 SUs on Frontera in total were used, which is equivalent to 156,000 CPU hours or 4.8 years of a normal desktop computer running continuously. “The extensive computational resources would not be feasible without the supercomputers at TACC.” Frontera is currently the most powerful academic supercomputer in the U.S. 


The collaboration with Buskey started in 2020 with a goal to better understand the impacts of the proposed deepening of the Corpus Christi Ship Channel. Valseth said that after the initial study on the fish larvae, he and Dawson met with UTMSI faculty and researchers to study other processes that could be impacted.

Valseth said he doesn’t often get to “physically see and touch the things we develop models of,” but last year he visited UTMSI and was given a boat tour of the area by Buskey and his team. “It was a very nice experience, since, as computational scientists, we usually work with computer models.” 

The next phase of the research focuses on effects of a deeper channel on the spread of a potential oil spill.

The present study was conceived of and initiated by scientists at the University of Texas Marine Science Institute and carried out by faculty and staff from the Oden Institute for Computational Engineering and Sciences of the University of Texas at Austin. The authors also gratefully acknowledge the computational resources provided by the Texas Advanced Computing Center and the Frontera supercomputer under allocations “ADCIRC” and “DMS21031.” The authors have no relevant financial or non-financial interests to disclose.


Eirik Valseth