How much rain could possibly fall during the most extreme storm — and how should engineers prepare critical infrastructure for it?

New research led by Dr. Gabriel Perez, assistant professor in the School of Civil and Environmental Engineering at Oklahoma State University, is challenging long-standing assumptions about how engineers answer that question, with potential implications for the design and safety evaluation of dams, spillways, hydropower facilities and other critical infrastructure.

The study, “Re-evaluating probable maximum precipitation estimates: sensitivity to transposition domains and storm rotation using modern datasets,” was published in the Journal of Hydrology. The research was conducted in collaboration with scientists at Oak Ridge National Laboratory and supported by the U.S. Department of Energy Office of Critical Minerals and Energy Innovation’s Hydropower and Hydrokinetic Office.

At the center of the research is probable maximum precipitation, or PMP, an estimate of the greatest amount of rainfall considered physically possible at a particular location. Engineers use PMP estimates when evaluating infrastructure where the consequences of failure could be significant.

“When I first started working on PMPs, one question kept coming to mind: How much of a PMP estimate is driven by the storm itself, and how much is driven by the methodological decisions made by the analyst?” Perez said.

Historically, PMP estimates have relied on methods established decades ago, including guidance from the National Weather Service’s Hydrometeorological Reports No. 51 and No. 52. Researchers found that several decisions within that process — including which historical storms are selected, where storms are moved, how they are intensified and whether they are rotated can significantly influence the resulting estimate.

Using the Red Rock watershed in Iowa as a case study, the research team analyzed high-resolution rainfall data from the National Oceanic and Atmospheric Administration’s Analysis of Record for Calibration dataset covering 2002-23.

The study found that PMP is not necessarily a single, fixed physical upper limit. Instead, estimates can be heavily influenced by assumptions made during the modeling process. The research team found that decisions involving storm sample size, geographic domain, climatological window and storm rotation all contributed to higher estimates.

Storm rotation proved particularly significant. Rotating historical storm patterns over the watershed increased the estimated PMP by more than 25% in the study.

The team also found that PMP values generated using modern gridded rainfall datasets could be substantially higher than the legacy estimate used in the original spillway design for Red Rock Dam.

The findings do not necessarily indicate that existing infrastructure is unsafe. Instead, they demonstrate the importance of understanding how methodological choices affect estimates used in infrastructure design and risk assessment.

As the engineering community moves toward more probabilistic approaches to estimating extreme precipitation, the researchers say incorporating physically meaningful storm characteristics and systematic uncertainty analysis will be important for developing more defensible methods.

“This study does not propose a new PMP methodology,” Perez said. “Instead, it takes a step back and systematically evaluates the sensitivity of the legacy PMP framework to key analyst-defined decisions, including storm selection, transposition domains, moisture maximization, and storm rotation.

“A major motivation for this study was the 2024 National Academies of Sciences, Engineering, and Medicine report on modernizing PMP estimation, which highlighted the need to move beyond legacy procedures by using modern observations, computational capabilities and physically based approaches. I hope this study helps contribute to that ongoing conversation by providing a clearer understanding of where uncertainty originates within the legacy framework.”

The study was authored by Perez along with Scott T. DeNeale and Shih-Chieh Kao of ORNL, and Felipe Quintero of the University of Iowa.

Read the full study in Journal of Hydrology: Re-evaluating probable maximum precipitation estimates: sensitivity to transposition domains and storm rotation using modern datasets