A levee breach is a structural failure that unleashes floodwaters—often with devastating consequences. The resulting flooding can be swift and overwhelming, destroying infrastructure and property, displacing residents, and putting lives at risk. For these reasons, predicting how levees fail is critical. Levee breach predictions provide essential data for flood modeling, emergency planning, and hazard mitigation projects.

Building a Levee Breach Database

I first started working on levee breaches with Dewberry in 2011 when I served as a member of a levee breach workgroup. FEMA tasked us with updating guidelines for levee breach scenarios. A big part of this was determining whether a levee is likely to fail. We set out to answer questions such as: If a levee fails, how big will the breach be? What kinds of levees tend to produce larger breaches? These questions are important for hydraulic modeling—you need to understand how large a breach might become before it reaches equilibrium, and how quickly it will reach that point.  

We prepared a report, but I wanted to take the data one step further. So, I continued the work on my own, gathering more data. Over time, I realized that most of the available data about levee breaches were anecdotal, incomplete, inaccurate, or very approximate. That realization marked the beginning of my long journey in search of reliable data.

I began building a database of levee breach cases and took on the role of an investigative journalist. If I wasn’t sure about the accuracy of certain measurements, I reached out to people on the ground—local sources, engineers, or the writers of reports—in an effort to verify information wherever possible. My goal was to collect as much accurate and credible data as I could for each case. The data I accumulated included breaches from both the U.S. and overseas.

Predictions for Preparedness

After compiling the database, I identified reliable data sets to develop empirical curves and equations to predict levee breach width and lateral erosion rates, the latter referring to the speed at which a breach develops laterally, including the time it takes to reach its full size.

My hope is that the equations can be integrated with hydrologic and hydraulic models that still use outdated assumptions for simulating levee breach scenarios.” Kaveh Zomorodi

I presented my research at the 2020 annual conference of the Association of Dam Safety Officials (ASDSO) and submitted a paper for the conference titled Empirical Equations for Levee Breach Parameters Based on Reliable International Data, which was shared on ResearchGate. The paper quickly began to generate interest within the professional community, including from the U.S. Army Corps of Engineers.

I further developed my paper and published a revised version under the same title in the Spring 2024 issue of the Journal of Dam Safety. This version introduces approximate equations for estimating peak discharge from a levee breach—a valuable planning tool for flood protection.

A key development occurred when I learned that my 2020 paper had been cited by researchers at the University of Bologna in Italy, who tested my equations for estimating breach parameters using data from 60 levee failures that occurred in Northern Italy in 2023. Their published findings demonstrated that the field verification was successful. The comparison indicated a good accordance between the predicted values and observed measurements, supporting the reliability of the equations. 

The field verification opens the door to broader use of the equations. My hope is that the equations can be integrated with hydrologic and hydraulic models that still use outdated assumptions for simulating levee breach scenarios. With more accurate predictions, we can strengthen our recovery and reconstruction efforts and, ultimately, improve community resilience following levee failures.