Scientists Predict Record-breaking 2018 Hurricane Season

Colorado State's team is predicting seven hurricanes in '18: slightly above-average, sure, but that would be much better than that outlier of 2017.

CSU, of course, is not the only one in the hurricane predictions game. University College of London has a group, Tropical Storm Risk, that predicts storms every year, too. (That group is calling for a decrease in Atlantic hurricane activity this year, about 15 percent lower than a long-term norm and about 25 percent lower than the norm for the past 10 years.)

The National Oceanic and Atmospheric Association's Climate Prediction Center is also forecasting a 45 percent chance of an above-normal season. Forecasters there predict a 70 percent likelihood of 11 to 17 named storms (winds of 39 mph or higher), five to nine of those becoming hurricanes (winds of 74 mph or higher), including two to four major hurricanes (Category 3, 4 or 5).

The Weather Company, an IBM unit that includes The Weather Channel, released its predictions as well, and is calling for 12 named storms during the 2018 season, including five hurricanes and two major hurricanes of Category 3 or higher.

James Belanger, the senior meteorological scientist at The Weather Company, explains how his team comes to its annual hurricane forecast. "We have some in-house analytics and capabilities that we look at. And then there's also kind of this human element," Belanger says. "A lot of times you might have a statistical model that provides you a baseline guidance. We look at a variety of statistical models and long-range weather forecasts — season forecasts that go out seven months — and we have access to some of those predictions. So, based on kind of a combination of that information, as well as what we're seeing in large-scale climate patterns, that's what we use to fine tune the numbers."

The differing ideas on the upcoming hurricane season demonstrate how complicated and inexact the science can be. Scientists all over the world gather information from more than 1,000 buoys, more than 3,000 ships, hundreds of weather balloons, several aircraft and a handful of satellites to track storms and measure things like water temperature, air pressure, winds and wind shear, just to name a few. They combine all that data with many other factors and compute it all through complex mathematical models (NOAA has a "Supercomputer System") to settle on its forecasts and predictions.

Ironically, some of the most important information the scientists rely on in predicting the Atlantic hurricane season isn't even taken from the Atlantic Ocean. Scientists lean heavily on measurements from El Niño and La Niña, a combined weather phenomenon in the central and east-central Pacific around the equator. (In the climate community, El Niño and La Niña also are known as ENSO, for El Niño/Southern Oscillation.)

El Niño is a water warming trend in the Pacific Ocean, especially those waters in the eastern Pacific, on the west side of Central America; La Niña is a cooling trend. The two are important because, generally, cooler water in the Pacific suggests warmer water in the Atlantic (and vice versa).

Monitoring the temperatures of the Pacific matters. Hurricanes are partly fueled by warmer water, so cooler water in the Pacific (a result of La Niña) generally means warmer water — and more hurricanes — in the Atlantic.

"When we're talking climate predictions, the Pacific tends to be the basin that has the most predictability," Belanger says. "We can make a prediction more skillfully in that region than anywhere else."

NOAA's prediction is based on no El Niño this season. "The outlook reflects our expectation of a weak or non-existent El Niño, near- or above-average sea-surface temperatures across the tropical Atlantic Ocean and Caribbean Sea, and average or weaker-than-average vertical wind shear in that same region,” Gerry Bell, Ph.D, lead seasonal hurricane forecaster with NOAA’s Climate Prediction Center, said in a statement.

Another well-known tracking tool that goes into the computer models for predicting hurricanes is known as the Atlantic Multidecadal Oscillation (AMO).

"El Niño and La Niña affect the hurricane season for a year at a time, while the AMO affects the season strength for decades (25 to 40) years at a time," Bell told an "Ask Me Anything" Reddit last year. "By predicting the combined impacts of these climate patterns, we can often predict what the hurricane season will bring. This is because they strongly control an entire set of key conditions such as vertical wind shear, trade winds, wind patterns coming off of Africa, ocean temperatures, and air pressure, all of which come together to make the season either more or less active."

The people and machines who predict hurricanes have a varying degree of success, depending on the year, of course. In May 2017, the CPC (which is a part of the National Weather Service, under the NOAA, all under the U.S. Department of Commerce) predicted an "above normal" season, with 11 to 17 named storms, five to nine hurricanes with two to four being major (Categories 3, 4 or 5) hurricanes for 2017. The '17 season produced 17 major storms, 10 hurricanes and six major hurricanes.