High above us lie "rivers in the sky"—narrow bands of moisture known as atmospheric rivers (ARs). These remarkable weather systems, spanning up to thousands of kilometers, transport immense amounts of water vapor across the midlatitudes, delivering crucial rainfall. While ARs are a vital water source, particularly for drought-prone regions like California, they also carry a destructive potential. Their sudden arrival can trigger catastrophic floods, landslides, and widespread damage, while their absence can increase wildfire risks.

In our recent study published in npj Climate and Atmospheric Science, we investigated the predictability of atmospheric rivers weeks to months in advance. Forecasting ARs on subseasonal-to-seasonal (S2S) timescales (10–30 days) is critical for preparedness and risk reduction, but it remains a challenging frontier in climate science. Our findings offer fresh insights into how large-scale climate patterns influence AR predictability and open up new opportunities for improving S2S forecasts.

How Did I Become Interested in Atmospheric Rivers?

My interest in atmospheric rivers grew during my early research on extreme weather events and their growing impact in a warming world. I noticed that ARs were often at the center of some of the most significant floods and drought relief events in regions like the western U.S. This duality fascinated me: how a single weather system could both replenish depleted water supplies and devastate entire communities. I was determined to explore whether it was possible to predict these "rivers in the sky" further in advance, giving communities and planners more time to prepare.

The opportunity to investigate this question came when I joined Princeton University and NOAA’s Geophysical Fluid Dynamics Laboratory (GFDL), where I had access to cutting-edge climate models and data. This collaboration laid the foundation for our study on S2S predictability of ARs.

Forecasting Atmospheric Rivers: A Behind-the-Scenes Look

At the heart of our research is the Seamless System for Prediction and Earth System Research (SPEAR) model, developed at NOAA’s GFDL. SPEAR is a seamless forecasting model designed to simulate and predict climate variability from days to decades. We used it to evaluate AR variability and predictability across the Northern Hemisphere, focusing on wintertime ARs, which are often the most intense and impactful.

One of the first challenges we faced was determining how well SPEAR could predict ARs beyond the standard weather forecast window of 1–2 weeks. While SPEAR performs well at short timescales, accurately predicting ARs in weeks 3–4 proved more difficult. However, we identified specific periods—what we call forecast windows of opportunity—when ARs become more predictable. These windows align with the activity of large-scale climate patterns like the El Niño-Southern Oscillation (ENSO), the Pacific-North American (PNA) pattern, and the Arctic Oscillation (AO), serving as dominant source of AR predictability at S2S timescales.

Further, our analysis revealed that La Niña (the negative phase of ENSO) enhances AR activity over the western U.S., improving forecast accuracy. Similarly, positive PNA and negative AO phases create favorable conditions for more reliable predictions. By understanding how these climate drivers influence ARs, we can significantly improve S2S forecasts.

Lessons Learned and Challenges Faced

One of the most rewarding aspects of this research was collaborating with a multidisciplinary team of scientists from climate modeling, meteorology, and data analysis backgrounds. Each team member brought a unique perspective that helped shape the study. However, the process was not without its challenges. Atmospheric rivers are highly dynamic, and capturing their variability at longer timescales required refining the model several times. Data availability and processing also posed obstacles, especially when working with large datasets spanning multiple decades.

Despite these challenges, the study was a breakthrough in understanding the predictability of ARs at S2S timescales. One key takeaway is that no single model or method is perfect. Improving forecasts requires integrating multiple models, continuously refining them, and incorporating new knowledge about climate patterns.

Why Does This Matter to the Community?

Accurately predicting atmospheric rivers weeks in advance has far-reaching implications. For water resource managers, it can mean better planning for water storage and flood control. For emergency responders, it provides crucial lead time to prepare for extreme weather events, potentially saving lives and reducing economic losses. Communities in wildfire-prone areas can also benefit by knowing when to expect extended dry periods.

On a broader scale, this research contributes to the ongoing effort to make our climate prediction systems more reliable and actionable. While there’s still much to learn, each discovery brings us closer to unlocking the secrets of these "rivers in the sky" and improving our understanding of the planet’s complex climate system.

What’s Next?

Looking ahead, my team and I are working on using high-resolution models to investigate how ocean mesoscale eddies influence AR activity. Ocean eddies play a crucial role in modulating heat and moisture transport in the atmosphere, which may have a significant impact on the formation and intensity of atmospheric rivers. By incorporating oceanic processes more accurately into climate models, we hope to gain deeper insights into AR behavior and further improve S2S forecast accuracy.

We are also exploring how AR activity connects to wildfire risks in California. Atmospheric rivers are not only responsible for heavy rainfall events but also for prolonged dry periods when they fail to make landfall. These dry periods leave vegetation parched and increase the likelihood of wildfires. Understanding this connection is critical for improving fire-weather predictions and supporting wildfire management strategies.

By continuing to unravel the mysteries of ARs, we aim to develop tools that empower communities to better prepare for both the blessings and risks that these "rivers in the sky" bring.