Bae et al. (2026) Spatial Covariability of Extreme Floods Over the Coterminous United States: Co‐Dependency Measures and Their Statistical Significance

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Water Resources Research
• Year: 2026
• Date: 2026-01-01
• Authors: Kichul Bae, Jeongwoo Hwang, A. Sankarasubramanian
• DOI: 10.1029/2025wr041262

Research Groups

Hydrology and Water Resources Research Groups; Climate Science Departments.

Short Summary

This study quantifies the spatial covariability of extreme floods across the coterminous United States using novel co-dependency measures, revealing that floods co-occur significantly more often than expected, with distinct physical drivers for different flood magnitudes and regions.

Objective

• To investigate the spatial covariability of extreme floods across the coterminous United States (CONUS) for return periods ranging from 2 to 100 years, and to associate this spatial dependence with underlying physical drivers.

Study Configuration

• Spatial Scale: Coterminous United States (CONUS), with analysis of co-occurrence within a 500 km radius.
• Temporal Scale: Return periods from 2 to 100 years, with analysis of annual and 7-day co-occurrence probabilities, considering seasonal variations (e.g., summer, winter).

Methodology and Data

• Models used: Three novel co-dependency measures: annual co-occurrence probability, 7-day co-occurrence probability, and Measures of Co-occurrence within a 500 km radius. Null distributions of spatially independent floods preserving seasonality were used for evaluation.
• Data sources: Not explicitly stated, but implied observational flood data across the coterminous United States.

Main Results

• Floods co-occur significantly more often than expected under spatial independence.
• Spatial dependence strengthens with increasing return periods; for 100-year floods, co-occurrence probability is approximately 19%, compared to 1% under independence.
• Snowmelt-driven basins show high flood dependence for smaller events (2–25 year return periods).
• Rainfall-driven coastal regions exhibit dominant dependence for extreme events (50–100 year return periods).
• Widespread extreme flood events are primarily driven by summer tropical storms on the East Coast and winter atmospheric rivers on the West Coast.

Contributions

• Introduction of three novel and effective co-dependency measures for quantifying flood spatial covariability across various spatial and temporal scales.
• Quantification of the extent to which extreme floods co-occur across the CONUS, highlighting stronger dependence for higher return periods.
• Identification of distinct physical drivers (snowmelt, tropical storms, atmospheric rivers) for flood spatial dependence based on flood magnitude and region.
• Provision of a framework that can inform regionally tailored, season-specific flood mitigation and emergency-response strategies.

Funding

• Not explicitly stated in the provided abstract.

Citation

@article{Bae2026Spatial,
  author = {Bae, Kichul and Hwang, Jeongwoo and Sankarasubramanian, A.},
  title = {Spatial Covariability of Extreme Floods Over the Coterminous United States: Co‐Dependency Measures and Their Statistical Significance},
  journal = {Water Resources Research},
  year = {2026},
  doi = {10.1029/2025wr041262},
  url = {https://doi.org/10.1029/2025wr041262}
}