For much of the United States east of the Rocky Mountains, the summer of 2012 brought a heat wave that smashed temperature records, triggered extensive crop failures, and helped fuel a powerful line of thunderstorms known as a derecho, which moved eastward out of Indiana to leave nearly 4 million customers without power as it eventually swept through the mid-Atlantic states. The heat wave and derecho accounted for more than 100 deaths.
Inspired by that summer's events, a team of researchers says it has found an unexpected pattern of sea-surface temperatures thousands of miles to the west that could give forecasters up to several weeks' warning that the central and eastern US may be due for a severe summer heat wave, beginning in a given week or even on a given day.
The study, published online Monday by the journal Nature Geoscience, highlights the growing interest in moving beyond forecasts that look at seasonal averages for temperature and precipitation across swaths of the US.
These have become more accurate over the past 20 years, researchers say. But they mask the extremes, which tend to inflict the most hardship on society.
Advanced forecasts of the likelihood of heat waves and when to expect them would be useful with or without global warming in the picture, researchers say. Such forecasts become more important as global warming intensifies such extremes.
“We certainly hope this can be used to increase the resilience of our society,” says Karen McKinnon, a scientist at the National Center for Atmospheric Research in Boulder, Colo., and the study's lead author.
Others not involved in the study say the results are intriguing, but they caution against expecting too much of the team's approach – at least at this stage of its work.
To be sure, some forms of forecasting that involve extremes, such as those associated with El Niño, have made a lot of progress, notes Mathew Barlow, a climate researcher at the University of Massachusetts at Lowell whose work focuses in part on understanding the climate system's natural variations over periods of days to months.
A few forecast offices regularly issue forecasts for extremes, especially heat waves. But the lead time for gleaning meaningful information can be relatively short.
Following the severe drought of 2011, for instance, researchers looked at how well the National Oceanic and Atmospheric Administration's National Centers for Environmental Prediction did in forecasting the event.
The researchers found that the forecasts correctly predicted a warmer-than-normal summer over the central Plains several months out. As the predicted onset of warmth neared, the forecast became more accurate about where and when temperatures were likely to soar. But the forecasts didn't capture the full intensity of the heat wave until about a week out.
Moreover, science journals have published “a large number of very promising-seeming forecasting methods that either ended up not being practical or not yielding useful operational results,” Dr. Barlow says.
Although the jury may still be out on how useful the team's work will be in day-to-day forecasting, it is providing valuable insights into some of the physical factors that drive summertime extremes, especially at mid-latitudes.
The team became interested in the 2012 event not only for its severity, but because it defied seasonal forecasts predicting good growing conditions for the Midwest, where farmers were hard hit, explains Peter Huybers, a climate researcher at Harvard University and the study's senior author.
Since sea-surface temperatures play a major role in climate and weather patterns, the team turned to daily data conditions in the Pacific Ocean north of the tropics to see what effect they might have. They examined data gathered between 1982 and 2015 and found a correlation between hot days at weather stations throughout the eastern US and patterns where large pools of warmer-than-normal water shared boundaries with large pools of cooler-than normal water in the central Pacific north of the tropics.
The sharper this pattern was, the higher the likelihood that the eastern half of the US would experience extreme heat.
As the team looked at the evolution of this pattern, they found that it could begin building as many as 50 days ahead of the onset of an extreme heat wave. At that lead time, the odds of a heat wave kicking in at the end of the forecast period ranged from 1 in 6 to 1 in 4. Thirty days out, the odds increased to 1 in 2 or better, depending on the pattern's distinctiveness.
They dubbed this pattern the Pacific Extreme Pattern (PEP), and used it to successfully “hindcast” the 2012 heat wave.
The researchers suggest that the PEP may help explain the appearance of two other features associated with the 2012 heat wave: a paucity of soil moisture in the weeks or months leading up to the heat wave, and the build-up of an atmospheric circulation pattern that in effect parks a vast high-pressure system over the center of the country.
The lack of soil moisture yields hotter near-surface temperatures than would be the case with moist soil. The high-pressure system allows more sunlight through largely cloudless skies to bake the surface.
How these factors interact remains a mystery.
It's unclear “if the ocean is causing the atmosphere to do what it does to make the heat wave or whether it's just sitting there watching this unfold as a passive entity,” says Martin Hoerling, a researcher at NOAA's Earth Systems Research Laboratory in Boulder, Colo., who focuses on the climate's natural variations and the role air-sea interactions play in them.
But if the team's identification of the PEP holds up, the researchers will have identified a potentially important feature of the ocean-atmosphere system that current forecast models fail to reproduce, UMass-Lowell's Barlow explains.
Although although a lead time of up to 50 days appears to be rooted in the PEP's natural behavior, it's unclear how useful that much advanced warming could be, researchers say.
Urban public-health officials, for instance, might set in motion their plans for helping residents cope with heat waves based on forecasts that give lead times of three to five days. Such forecasts could trigger a review of cooling shelters ready to accept people whose homes have no air conditioning. Preparations might also include public-service ads reminding people to check on the elderly or neighbors with small children to see if they need to move to shelters.
Electric utilities also would need to gear up for the extra demand they would experience during heat waves.
Today, “as far as we can tell, there are no long-lead-time forecasts for summer extremes, so it's hard to know exactly how long would be helpful, because it's not really available at this point,” says Dr. McKinnon.
“But we believe our lead times should be long enough for some adjustments,” she adds. If utilities knew there was a high probability of heat waves for a coming summer, they could defer maintenance on some power plants to ensure they remain available to meet the demand. Or utilities could arrange to bring extra plants on line.
Because skillful long-term forecasts for bouts of extreme weather have yet to appear, “no one is thinking about how to use them,” says Barlow. As such forecasts evolve, he adds, “there'll be lots more discussion” on possible uses.