Meteorologists say they’ve found a pattern in Pacific Ocean sea surface temperatures that could help authorities prepare for heat waves in the eastern United States up to 50 days in advance.
Now that the pattern has been found, forecasters will start keeping track of the heat wave indicators in May. But don’t expect the 50-day forecast to show up in the nightly weather report.
“Most seasonal predictions, including this one, are probabilistic rather than deterministic,” lead author Karen McKinnon, a climate scientist at the National Center for Atmospheric Research in Colorado, explained in an email. “For example, we can predict an increase in the odds in favor of having a hot day in the Eastern U.S. from about 1 in 6, to 1 in 2, at lead times of 40 days if the Pacific Extreme Pattern is particularly strong.”
She said the indicators are most likely to come into play during preparations for the peak of the summer.
“For example, city leaders could ensure they have sufficient cooling rooms for the elderly or those without air conditioning; farmers could alter their management tactics to prevent crop loss; businesses could be prepared for increased demand of air conditioners and fans; and utilities could ensure they have sufficient power options available to bring online quickly in case of a spike in demand,” she said.
The research was published today by Nature Geoscience. One of the authors, Andy Rhines, is a climate scientist at the University of Washington.
The Pacific Extreme Pattern was identified in a wide region of the Pacific extending north from Hawaii: If sea surface temperatures in the southern part are unusually warm, and temperatures in the northeastern part are unusually cold, there’s a increased chance of having an Eastern heat wave, looking as far out as 50 days out. The farther out the prediction is made, the less accurate it becomes.
McKinnon and her colleagues developed the model by matching up Pacific surface temperature patterns with high-temperature summer days in the eastern U.S. between 1982 and 2015. Once they identified the model, they tested it by seeing how well it would have done during the extraordinarily hot summer of 2012. They found that their model was more accurate than the standard forecast model for predicting heat waves 30 to 40 days in advance.
Why are temperatures in the Pacific correlated with hot days that take place weeks later, thousands of miles to the east? It’s not completely clear, but the leading hypothesis is that Pacific temperature shifts affect atmospheric circulation downstream. Those shifts would influence the flow across the United States to create the persistent high-pressure, low-precipitation weather pattern that’s associated with heat waves.
The researchers said the Pacific Extreme Pattern marks “an important advance over current seasonal forecast models that tend to underpredict the probability of extremes.” But there’s lots more work to be done.
For example, the study doesn’t account for the potential effects of climate change. McKinnon suggested that the heat waves predicted by the newly identified pattern wouldn’t become more frequent, but would become more intense. “It’s important to note, however, that this is simply informed speculation,” she said.
McKinnon also noted that the findings don’t say much directly about weather in the western United States.
“However, the atmospheric circulation anomaly associated with the Eastern U.S. hot weather also includes a relatively low-pressure system around the Gulf of Alaska, which could induce anomalous offshore flow into the Pacific Northwest,” she wrote in her email. “Additional analysis would be necessary to conclude whether or not this occurs, and what the impact would be on the ground in the West.”
Similar models could also be developed for long-range predictions of winter blasts like the ones that hit the eastern U.S. in 2014 and 2015.
In addition to McKinnon and Rhines, the authors of “Long-lead Predictions of Eastern United States Hot Days From Pacific Sea Surface Temperatures” include Martin Tingley and Peter Huybers.