CW3E Publication Notice

Using Model Climate to Improve Situational Awareness Ahead of Atmospheric River-Related Landslide Events in Southeast Alaska

September 30, 2026

Atmospheric rivers (ARs) are an important source of precipitation in Southeast Alaska, but they can also trigger widespread flooding and landslides (example shallow landslide in Fig. 1). A new study from CW3E and the National Weather Service (NWS) Juneau Forecast Office introduces an operational tool designed to help forecasters identify when an AR is accompanied by meteorological conditions that can increase the potential for these hazards.

Figure 1.Example of a shallow landslide triggered by precipitation during an atmospheric river that followed a snow event on January 1, 2019 near Kasaan, Alaska. Photo credit: D. Nash, July 2025

The paper, “Using Model Climate to Improve Situational Awareness Ahead of Atmospheric River-Related Landslide Events in Southeast Alaska” was recently published in Journal of Operational Meteorology – Forecaster’s Toolbox. The study presents the AR Hazard Index, a forecast tool developed in close partnership with NWS Juneau, that uses model climate (M-Climate) percentile ranks from the Global Ensemble Forecast System Reforecast v12 (GEFSv12) to provide historical context for forecasts.

The AR Hazard Index combines several meteorological ingredients relevant to landslide and flood hazards, including integrated vapor transport (IVT) magnitude, duration, and direction, low-level wind magnitude and direction, freezing level, and quantitative precipitation forecasts (QPF). Rather than considering precipitation or AR strength in isolation, the tool identifies when multiple forecast variables are unusually extreme compared with the model’s historical climate.

The results show a strong relationship between these combined extreme conditions and observed landslide activity. Approximately 90% of landslide days had extreme low-level wind, precipitation, and IVT conditions, compared with approximately 30% of non-landslide days. Here, extreme conditions are defined as all three variables reaching or exceeding their 95th percentile. For example, figure 2 shows an example of what the tool would have looked like for 2 December 2020 at a 72-h lead when a warm, windy, long duration and intense AR resulted in widespread landsliding and flooding.

Figure 2.What the model climate AR Hazard Index tool looked like for 00Z 2 December 2020 (F-072) during a rain-on-snow atmospheric river event that triggered widespread landslides and floods across Southeast Alaska.

~90% of landslide days had combined extreme conditions compared with ~30% of non-landslide days, landslide days were much more likely to have extreme low-level wind, precipitation, and IVT, with all three variables at or above their 95th percentile.

The study also found that 75% of the 120 landslide days had forecasted IVT, low-level wind, and QPF at or above the 98th percentile, together with an AR Index value of at least 2.38 at a 24-hour lead time. Similar results were found at forecast lead times out to 96 hours, indicating that the M-Climate percentile ranks provide useful and consistent historical context at longer lead times.

These results support the use of M-Climate as a tool for improving situational awareness ahead of potentially hazardous AR events. By placing a forecast into historical context, the AR Hazard Index can help forecasters recognize when an upcoming storm contains an unusual combination of conditions associated with landslide and flood impacts.

The tool was developed through a close research-and-operations partnership between CW3E and the NWS Juneau Forecast Office and is now available for operational use. This collaboration provides a direct pathway for applying CW3E research on atmospheric rivers, extreme precipitation, and geohazards to operational forecasting in a region where complex terrain and limited observations make predicting impacts particularly challenging.

“The M-Climate tool helps forecasters gain confidence on potential impacts from atmospheric rivers across Southeast Alaska as it is a combination of components of ARs that lead to impacts of flooding and landslides and is calibrated to a specific time of year.” – Aaron Jacobs (NWS Juneau)

The AR Hazard Index supports a key priority in CW3E’s Strategic Plan: advancing atmospheric river and extreme precipitation research, prediction, and applications. More broadly, the work demonstrates how model climate information can connect atmospheric river forecasts to the potential impacts of extreme weather on communities and landscapes.

Explore the tools:

Citation:

Nash, D., Rutz, J., Jacobs, A., and Kawzenuk, B. (2026). Using Model Climate to Improve Situational Awareness Ahead of Atmospheric River-Related Landslide Events in Southeast Alaska. Journal of Operational Meteorology, 14 (18) https://doi.org/10.15191/nwajom.2026.1418