U.S. West flood risk increases sharply with wet soils during atmospheric river storms

A peer-reviewed study published on June 4, examined the influence of antecedent soil moisture on flooding caused by atmospheric river storms across the U.S. West Coast.

Using a dataset of more than 43 000 events in 122 watersheds spanning the years 1980–2023, the research shows that flood magnitude depends not only on storm intensity, but also on the condition of the soil before the storm arrives.

When soils were already wet, observed flood peaks during atmospheric river events were, on average, 2–4.5 times higher than for similar storms where the land was drier.

The increase in flood response was found to be non-linear, after a critical threshold of soil moisture is crossed, any additional rainfall leads to an abrupt jump in runoff and streamflow, rather than a gradual rise.

The study found that even moderate atmospheric river storms can cause major floods if they strike when soils are saturated. Conversely, some of the most powerful atmospheric river events passed with little flood impact in areas where soils were dry at the onset.

The analysis identified specific soil moisture thresholds for each watershed. When antecedent soil moisture exceeded these values, the watersheds were especially vulnerable to large floods.

This effect was most pronounced in arid and semi-arid regions like California and southwestern Oregon, where soils are shallow, often rich in clay, and where there is less rainfall and higher evaporation.

Soils in these areas tend to experience more variability in moisture content. When storms arrived after soils had surpassed their storage capacity, the risk of flooding increased sharply.

In contrast, humid regions such as western Washington, the interior Cascades, and the Sierra Nevada have deeper soils and more consistent snowpack, which buffer the effect of soil wetness. In these places, soils are often already saturated or covered by snow, so extra rainfall adds less to flood risk.

Flood forecasting and risk assessment can be improved by incorporating real-time soil moisture measurements with weather models, particularly in high-risk, variable-moisture watersheds.

The study suggests prioritizing installation of soil moisture sensors in basins with low storage and high antecedent moisture variability to allow for timely flood risk assessments.

The researchers recommend moving away from generalized storm-flood models in favor of approaches that account for the specific antecedent soil moisture thresholds of each watershed.

In agricultural regions like California’s Central Valley, real-time soil moisture data could help distinguish between atmospheric river-driven flooding and flooding from irrigation, and also support groundwater recharge planning.

At present, soil moisture is measured at relatively few points, mainly by networks like the USDA’s SNOTEL, so coverage is sparse compared to rainfall monitoring. Soil moisture can also vary greatly within a single watershed, making expanded and more granular monitoring essential for early warning as atmospheric rivers become more frequent and intense under climate change.

Atmospheric rivers are responsible for most major floods along the U.S. West Coast, causing more than USD 1.1 billion in damage annually, but they also provide much-needed precipitation.

---

Also read: Pacific clues improve atmospheric river forecasts

---

This research brings together atmospheric science and hydrology, making it clear that the condition of the land surface, especially soil moisture, determines whether an atmospheric river event results in beneficial rain or catastrophic flooding.

By refining risk forecasts and expanding sensor networks, local agencies and emergency managers can better anticipate and prepare for flood events as atmospheric river impacts grow in a changing climate.

The study’s findings are based on decades of USGS streamflow records, high-resolution soil moisture modeling, and detailed tracking of atmospheric river storms. The researchers measured soil moisture as a proportion of the soil’s storage capacity in the upper two meters, averaged across each watershed.

They compared streamflow peaks during AR events under wet and dry conditions and quantified how much of the response was explained by soil moisture rather than precipitation alone.

In most watersheds, a clear moisture threshold was found. Once this threshold was exceeded, average flood peaks more than doubled, and in some cases increased by more than tenfold compared to dry events.
The strongest increases were seen in central and southern California, while the lowest were in higher elevations with more snowpack.

The study groups watersheds into those with a strong or weak response to soil moisture, with the highest sensitivities in California and Oregon coastal basins and the weakest in Washington and the mountain interior.

The results point to the need for targeted monitoring and soil moisture-informed flood forecasting.

References:

¹ Wet Antecedent Soil Moisture Increases Atmospheric River Streamflow Magnitudes Nonlinearly – Mariana J. Webb, Christine M. Albano et al. – Journal of Hydrometeorology – June 4, 2025 – https://doi.org/10.1175/JHM-D-24-0078.1