This project aims to compare publicly available downscaled projections at the continental U.S. scale and provide guidance to decision-makers and stakeholders.

There is a growing consensus on increasesincreases in the frequency and intensity of rainfall extremes that will continue throughout the 21st-century. We need adaptation actions to increase the resilience of new and existing urban stormwater infrastructure systems under a changing climate. Examples of these systems are culverts, drainage inlets, etc.

Many of the existing systems were designed based on historical records to satisfy a specific performance threshold that is established by local regulations or State Department of Transportation. We expect that the performance of these systems will continue throughout its service life, but the problem is that they will be exposed to a future with more frequent and intense storms. We can evaluate future climate change impacts on existing infrastructure by using local and regional projections from climate models.

Several efforts to provide higher spatial resolution projections either through statistical or dynamical downscaling have resulted in multiple publicly available bias-corrected climate projections datasets. Previous studies frequently relied on a single dataset for assessing future infrastructure performance, without accounting for differences in future rainfall extremes in the available datasets. These variations could result in different adaptation decisions by stakeholders, and it is this information that we want to integrate into the design decision-making of stormwater structure. We characterized the differences between 4 downscaled climate projection datasets (NA-CORDEX, MACA, LOCA and BCCAv2), and identified the predominant source of uncertainty of daily rainfall extremes defined by return periods ranging from 2- to 100-year across the United States. We believe that the results enhance our understanding of stormwater infrastructure vulnerability to future extreme rainfall under a broad set of plausible scenarios and enable stakeholders to undertake robust adaptation decisions.

This work has been published in Geophysical Research Letters Lopez-Cantu, T., Prein, A. F., & Samaras, C. (2020). Uncertainties in Future U.S. Extreme Precipitation From Downscaled Climate Projections Geophysical Research Letterss. 47(9) e2019GL086797


  • Large differences in magnitude and spatial patterns of extreme precipitation changes exist between datasets.
  • High-end extremes (i.e., 100-year event) increase 10-50% faster than low-end extremes (i.e., 2-year event) at the continental scale.
  • While short record lengths dominate uncertainties for changes in high-end extremes, dataset differences largely contribute to uncertainties in more frequent extremes.
  • Uncertainties across datasets need to be included in climate resilience decisions and impacts analyses to increase robustness.

Read the published paper for free here.