Posted: August 25, 2025

Coastal flooding in the United States has increased over the past several decades, especially at high tide. More flooding means more damage to property and infrastructure.

While predicting coastal floods has typically been limited to a few days in advance, a team of Physical Sciences Laboratory and National Ocean Service researchers led by the PSL’s John Albers have demonstrated for the first time that operational forecast systems can be used to predict unusual coastal water levels several weeks in advance.

The findings of their study published in the European Geosciences Union’s Ocean Science could be used to improve flood risk forecasting, giving coastal communities more time to prepare.

The approach

• Researchers studied subseasonal retrospective forecasts (forecasts of past weather conditions run through modern models) from two forecast systems for the years 2000 to 2019 to see how well they predicted the non-tidal residual component of coastal water levels. Non-tidal residual (NTR) sea level is the “base” sea level that does not include the regular rise and fall of the tides. NTR can still fluctuate based on non-tidal factors such as wind, air pressure, currents, and storm surge.
• The European Centre for Medium-Range Weather Forecasts Integrated Forecasting System (IFS) and the Centre National de Recheres Meteorologiques (CNRM) climate model were used. These two systems were selected because they both include ocean height data - important in sea level forecast skill - in their initial conditions.
• The predictions made by these two forecast systems were compared with real-world observations from U.S. coastal tide gauges that had captured at least 10 years of data. Tide gauges measure the sea level at a specific coastal location over time. Station counts were 47 on the East Coast, 32 on the Gulf of America (formerly known as the Gulf of America), 35 on the West Coast, and 23 in Alaska.
• As part of this evaluation of forecast skill, researchers also made improvements to the forecasts by adding air pressure effects and adjusting for land that’s slowly rising or sinking.
• The resulting forecasts were examined for accuracy and usefulness using standard forecast skill assessment methods.

The results

• Both models were compared to ‘damped persistence’, which is the current standard used by the National Ocean Service to forecast non-tidal residual anomalies. The damped persistence method assumes current conditions will continue into the future but with gradually decreasing influence over time.
• Both the IFS and CNRM predictions were found to be able to skillfully predict unusual sea level changes better than damped persistence 2-3 weeks out. Only the IFS was more skillful beyond 3 weeks.
• Accounting for air pressure and vertical land motion greatly improved prediction accuracy in many geographic areas.
• Forecast skill was better in the winter when coastal flooding tends to be worse.
• Not all regions were as skillfully predicted as others. California, parts of Alaska, and parts of the Gulf coast had a high degree of skill. The East Coast was less skillful for the Midatlantic states compared to New England and the southern Atlantic coast.

The impact and what's next

• The study demonstrated for the first time that current operational forecast models can be used to predict the risk of coastal flooding for many US coastal locations up to six weeks in advance with sufficient skill.
• The findings of this study form the basis for improved forecast guidance of high tide flooding predictions on subseasonal (2-6 weeks) timescales. Improved forecasts will give coastal communities more time to prepare, reducing the risk to lives and property.
• The results also suggest that future coastal flooding forecast model development would benefit from including air pressure effects due to its significant impact on the forecast skill.
• The next step is to test and validate the new ocean-coupled NOAA Global Ensemble Forecast System version 13 (GEFSv13) for use as NOAA’s forecast model to meet the needs of the high tide flood outlook.
Read the journal article

---

Back to PSL News   |   Signup for PSL email updates