Liao et al. (2026) Estimating probable maximum precipitation for the continental United States

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2026
• Date: 2026-01-13
• Authors: Mochi Liao, Ana Grazielle Soeiro Barros
• DOI: 10.1016/j.ejrh.2026.103122

Research Groups

• Civil and Environmental Engineering, University of Illinois Urbana-Champaign, Urbana, IL, United States

Short Summary

This study estimates Probable Maximum Precipitation (PMP) for the continental United States using multifractal analysis with high-resolution precipitation datasets, revealing strong spatial alignment between PMP, multifractal parameters, topography, and weather regimes. The research demonstrates that AORC-derived PMP estimates align well with historical records for a 1000-year return period, highlighting significant differences between model-based and gauge-corrected products and the sensitivity of PMP estimates to data record length and extreme events.

Objective

• To estimate Probable Maximum Precipitation (PMP) using multifractal analysis for the continental United States (CONUS).
• To evaluate the impact of recent extreme events on PMP estimation and associated return periods using high-resolution precipitation datasets (ERA5L and AORC).

Study Configuration

• Spatial Scale: Entire continental United States (CONUS).
• Temporal Scale: Hourly precipitation data over periods of up to 75 years (1950–2024 for ERA5L) and 47 years (1979–2024 for AORC), focusing on 24-hour PMP and return periods up to one million years.

Methodology and Data

• Models used: Multifractal analysis based on a multiplicative random cascade model using a Lévy-stable distribution.
• Data sources:
  • ERA5-Land (ERA5L) reanalysis at 9 km spatial resolution.
  • Analysis of Record for Calibration (AORC) multi-sensor gauge-corrected reanalysis at 4 km spatial resolution.

Main Results

• A strong spatial alignment was found between extreme precipitation, multifractal parameters (α and C1), topography, and regional weather regimes.
• A large magnitude gap exists in estimated PMP between model-based (ERA5L) and multi-sensor gauge-corrected (AORC) precipitation products.
• For a 24-hour duration, PMP estimates for return periods of one thousand and one million years are approximately 400 mm and 2000 mm, respectively, using ERA5L, and 800 mm and 6000 mm, respectively, using AORC.
• Precipitation accumulations from recent extreme events align well with PMP estimates derived from multifractal analysis using AORC for a 1000-year return period.
• Multifractal parameters are sensitive to the length of data records, with a minimum of 30–40 years of hourly precipitation data required for stable estimates.
• A single extreme event (e.g., Hurricane Helene) can significantly impact multifractal parameter estimates, causing a change of approximately 30% in the α parameter.
• PMP results indicate that extreme events have become more common since 1979.

Contributions

• Provides an alternative perspective for analyzing precipitation extremes through scaling analysis, contrasting with traditional statistical methods.
• Leverages the multifractal framework to estimate PMP across CONUS for desired return periods, consistent with engineering design risk analysis, using recently developed high-resolution datasets.
• Demonstrates that multifractal parameters exhibit spatial patterns aligned with regional topography and climate regimes, indicating the multidimensional nature of PMP spatial features.
• Highlights the critical sensitivity of multifractal parameters to data record length and the influence of individual extreme events, underscoring the importance of continuous, long-term, high-resolution precipitation measurements for robust PMP estimation.
• Illustrates the importance of continuously monitoring PMP estimation in a changing climate and urges future studies to analyze the impact of climate variability on extreme rainfall characteristics.

Funding

• Funds provided by the Grainger College of Engineering.

Citation

@article{Liao2026Estimating,
  author = {Liao, Mochi and Barros, Ana Grazielle Soeiro},
  title = {Estimating probable maximum precipitation for the continental United States},
  journal = {Journal of Hydrology Regional Studies},
  year = {2026},
  doi = {10.1016/j.ejrh.2026.103122},
  url = {https://doi.org/10.1016/j.ejrh.2026.103122}
}