The University of Arizona

Early Insights into the effects of Climate Change on Atmospheric Rivers

December 9, 2011
Author Profile: 

Fierce winds loaded with moisture blasted into the Southwest on December 18, 2010, leaving a drenched path in its wake. Rain and snow fell in epic proportions from southern California to southern Colorado. In St. George, Utah, 14 inches of rain drenched the landscape over several days. Parts of northwest Arizona recorded as much as 6 inches between December 18–23, 2010, single-handedly beating back drought from smothering the area as it did for most of the Southwest during the dry winter that followed. A similar dramatic winter storm occurred earlier last year in mid-January when a soggy six-day span delivered as much as 5 inches of rain to the Phoenix area, or about 34 percent of the total annual precipitation.

The bearer of these events was a recently discovered phenomenon called ”atmospheric rivers.” ARs, as they are known to scientists, often deliver extreme precipitation and have caused nearly all the largest floods on record in California. The state spends about $400 million each year repairing flood damage, and about 90 percent of that is because of ARs. The high price tag, not to mention the lives they disrupt, makes accurately forecasting ARs (see recent blog) and assessing potential changes in their intensity and frequency critical targets for research. Many advances in scientific understanding are yet to come, but initial research suggests that global warming will alter the character of ARs.

Photo Credit: Mindy Butterworth

Mike Dettinger, a research hydrologist at the US Geological Survey at the Scripps Institution of Oceanography in Southern California, published one of the first climate-change impacts studies focused on AR events in the June 2011 Journal of the American Water Resources Association. His results suggest that some winters will experience a larger number of AR events, with the largest events becoming more intense.

Dettinger used data from seven global climate models used in the Fourth Assessment Report (AR4) of the Intergovernmental Panel on Climate Change (IPCC) that were driven by the A2 greenhouse-gas (GHG) emission scenario. In the A2 scenario, GHG emissions rise unabated throughout the 21st century and are nearly five times more than they were in 1990; global temperature rises about 7 degrees Fahrenheit. It’s a scenario that should be avoided to prevent dangerous consequences. Dettinger chose this path because it provides clear indications of human-caused change amidst the natural variability of ARs. His analysis can be thought of as a realistic upper limit to changes in these events given current understanding.

Dettinger compared the frequency and intensity of ARs during 1961–2000, 2046–2065, and 2081–2100 periods. Modeled AR events for the historical period matched reasonably well with observations, providing credibility that the models can detect ARs. The future time slices were chosen for periods where daily data was available. In the models, ARs were identified when near-surface wind speeds averaged 10 meters per second (about 22 miles per hour) and the amount of water vapor integrated over each vertical sliver of the water column was greater than 2.5 cm (1 inch), a definition currently used in AR forecasting. Precipitation was not analyzed because computing power prevents the incorporation of realistic topography in general circulation models (GCM). Since mountains play a central role in instigating precipitation, the models would underestimate the quantity and rates of rain and snow. The lack of realistic topography confines the analysis to model grid cells just off the coast of California.

The intensity of ARs is determined by combined effect of the amount of water vapor and the speed at which it is carried by the winds. Dettinger’s analysis suggests that while future ARs will ferry more water vapor, winds will slacken. Both of these results have physical explanations. Warmer air can hold more water, a fact known since about 1834. Scientists have also hypothesized that winds speeds decrease because the temperature gradient between the poles and topics—which is a control on the intensity of atmospheric circulation—declines as higher latitudes warm faster than lower ones.

The weaker winds almost compensates for the higher water vapor so GCMs suggest that the average intensity of the storms may not change all that much. However, extreme AR events become more intense. Most of the models also simulate an increase in the number of ARs. By the end of the century, their average number increases by about 2.5 per year.  Most models also simulate more winters with a greater number of AR storms and fewer winters with a low number of events.

Increases in global temperature also means AR storms will be warmer. The models project AR events occurring around mid-century will be about 1.8 degrees Fahrenheit warmer than current storms. This will elevate the snowlines in the mountains, increasing the area receiving rain instead of snow, and, all else being equal, will cause larger floods.

Dettinger cautions that his results are a first crack at assessing future changes in ARs and their attendant impacts. Analysis of more GHG emission scenarios will help refine estimates, as will improvements in models. For Arizona, more on-the-ground research is needed to quantify how ARs have affected the region, as well as related model queries to explore implications. What we know now is that although ARs predominantly strike the West Coast, a few do stream into the Southwest (see recent blog). And for those that do squat over the region, like the two that struck Arizona in 2010, continued global warming may boost their intensity.


*Dettinger, Michael, 2011. Climate Change, Atmospheric Rivers, and Floods in California – A Multimodel Analysis of Storm Frequency and Magnitude Changes. Journal of the American Water Resources Association (JAWRA) 47(3):514-523. DOI: 10.1111/j.1752-1688.2011.00546.x