Ocean Temps May Predict Heat Waves 50 Days Out

By Christina Phillis

Transport yourself back to the summer of 2012. The heat wave that occurred was one of the hottest and longest-lasting in United States history. Yet the forecast issued by the National Oceanic and Atmospheric Administration for that summer predicted normal temperatures in the Midwest and Northeast U.S. This forecast was wrong not once but three times that summer with major heat waves occurring in late June, mid-July and late July.

A team of researchers set out to identify what triggers heat waves like the ones witnessed that summer so they could be more accurately predicted in the future. According to their findings, which will be published in the journal Nature Geoscience, patterns in Pacific Ocean temperatures can help to predict heat waves up to 50 days in advance. The specific pattern to look for is warmer-than-average ocean water butting up against colder-than-average ocean water. Depending on how well formed the pattern is the odds that extreme heat will follow can more than triple.

Study Method and Results

Postdoctoral researcher at the National Center for Atmospheric Research (NCAR) and lead author of the study Karen McKinnon and her colleagues started their research by dividing the U.S. into areas that experience extreme heat at the same time. The largest block, home to both major cities and farmland, stretches across much of the Midwest and up the East Coast. It did not take them long to find a pattern between extreme heat in the Eastern U.S. and global sea surface temperature anomalies. One such pattern was found above 20 degrees north latitude in the middle of the Pacific Ocean, roughly the same latitude as the United States. The phenomenon, which scientists have named the Pacific Extreme Pattern, appears not only when it is already hot in the Eastern U.S., but in advance of a heat wave as well.

To test the effectiveness of this prediction method, the team of researchers looked at temperature data collected from 1,613 weather stations across the Eastern U.S. between 1982 and 2015, as well as sea surface temperatures for the same time period. They focused their analysis on the hottest days of the year for that region, which usually occur between June 24 and August 22. When the warmest 5% of weather stations in that region reported temperatures 6.5 degrees Celsius (11.7 degrees Fahrenheit) above average, those occurrences were defined as extreme heat.

By “hindcasting” each year in the dataset, researchers attempted to see if they could retrospectively predict extreme heat events during the summer of 2012. Using the ocean pattern scientists were able to predict an increase in the odds ― from about 1 in 6 to 1 in 4 ― that a heat wave would strike 50 days in advance. The odds that a heat wave would occur on a particular day increased to better than 1 in 2 for a well-formed pattern when they looked 30 days out.

Using this method also allowed researchers to predict extreme heat for individual weather stations, located mainly in the middle of the country. The reasoning behind their success in predicting heat waves in this region, explained McKinnon, is that the domes of hot air associated with heat waves tend to be centered on that area of the country.

Since the typical seasonal forecast is only able to predict general trends for an entire summer as opposed to individual days, this new method has the potential to improve upon traditional forecasting measures. “It’s been very hard to get accurate, long lead-time prediction,” Noah Diffenbaugh, a climate scientist at Stanford University who wasn’t involved with the research, told Scientific American. “This paper is kind of extending our potential to have this longer time-scale predictability.”

Societal Effects of Extreme Heat

As more and more heat is trapped by greenhouse gases and the warming of the planet continues, it is predicted that heat waves will increase in occurrence and intensity. The future impact this will have on agriculture, infrastructure and vulnerable populations is becoming more concerning. The ability to make long-term weather predictions will be that much more important, giving farmers, cities and utilities more time to prepare and make decisions to manage the risk.

The heat wave that occurred in the summer of 2012 cost an estimated $31.5 billion due to the drought it caused and the intensity of the heat. A Time article that ran in the summer of 2012 described more specifically some of the potential effects of high heat and droughts. According to that article, some of the effects include an increase in the price of food as animals and crops die out, hazardous travel conditions due to buckling roads, and large strains being placed on healthcare systems that have to deal with an increase in heat-related illnesses. A study conducted on a 2006 California heat wave reported an increase of 16,000 emergency room visits and an increase of 1,200 hospitalizations during that period.

Next Steps

It’s still not clear to researchers why sea surface temperatures in the Pacific Ocean predict heat in the Eastern U.S. One possibility is that sea surface temperatures initiate weather patterns that cause the heat. Another possibility is that one does not cause the other and they are just two different results of the same phenomenon.

Learning how the two are connected using sophisticated computer models will be the next step for McKinnon and colleagues at the NCAR. The results of this study also could lead to the possibility of predicting other weather events further in advance, such as cooler-than-average days and extreme rainfalls.

“The results suggest that the state of the mid-latitude ocean may be a previously overlooked source of predictability for summer weather,” McKinnon told Science Daily.

As soon as this year, the team of researchers from the NCAR will try using ocean patterns to make forecasts in real time. Although the forecast track record using this new method may have been good over several years, there are a variety of other factors that affect the weather. Michael Ventrice, operational meteorologist with The Weather Company, told Scientific American that this is an intriguing possibility and a good case study in understanding whether or not this prediction method truly works.