Before the Atlantic hurricane season began in June, forecasters had for months warned that it would be an especially lively one. Conditions, they warned, were ideal for stoking storms in the Atlantic: Sea surface temperatures were high, for example, and wind shear was low. But the ferocity of this season's storms still took many by surprise.
Hurricane Harvey shattered rainfall records in Texas and Hurricane Irma left dozens dead and millions without power across the Caribbean and mainland United States. The two monster storms caused up to $200 billion in damage to Texas and Florida. As Irma carved its path of death and destruction, Hurricane Jose spiraled through the Atlantic and intermittently threatening to make landfall in the Leeward Islands and Florida, and Hurricane Katia slammed into Mexico, which was still reeling from a devastating earthquake. The Atlantic hurricane season doesn't end until November 30th.
Pacific Standard spoke with Colorado State University's Phil Klotzbach, an expert on seasonal hurricane forecasts in the Atlantic, about this hurricane season, the record-setting storms we've seen so far, and what might be yet to come.
How does seasonal forecasting work?
When we forecast the season, we're trying to look at the large-scale environment and say whether or not it looks more or less conducive for storms to form, and then predict numbers based off of that. As these thunderstorm complexes move off of Africa, if they move into a more conducive environment, that just kind of loads the dice, and makes hurricanes more likely to form. In early August, when we put out our final forecast, we said it looks like a pretty active season. Now you can't say in early August that Hurricane Irma is going to form, and it's going to track due west and plow through a bunch of islands, and then come up and scare the sock out of everyone in Florida. But what you can say is, "Hey, the environment looks conducive for storms to form."
Has this season been living up to your expectations so far?
It's definitely a little more active than I thought it was going to be. This season was kind of plodding along near normal until, honestly, even through Harvey. Irma was just such a long-lived major hurricane—there's not a lot of Irmas. There's not a lot of storms that last that long and are that intense. We've seen them, but you can go many years without seeing a storm like Irma. I mean, this was a buzzsaw through the Caribbean. Even now, it's been over land for almost a day and its still bringing 50-, 60-, 70-mph wind gusts, so it's still just very slowing spinning it's way down as it goes into Georgia. But unlike Harvey, at least this storm has the decency to move.
If the storm hadn't stalled, we wouldn't have been talking about Harvey for more than a day, it would have been almost a non-event. It just sat, and it sat close enough to the Gulf of Mexico that it could just keep pulling in all this moisture, and then it dropped all that rain in Houston. Irma was known mainly for its intensity. We see Category 5s obviously, there's plenty of those in the Atlantic, but to see one as strong as Irma, for as long as Irma was that strong, was quite notable.
How do you measure intensity?
There's an integrated measure that NOAA uses that's called Accumulated Cyclone Energy. Basically you take the wind speed every six hours where you have a tropical storm or a hurricane—any named storm—and then take that wind speed and you square it. This year, Irma has generated more of this Accumulated Cyclone Energy than 18 of the previous 51 years, which is the era that we've had satellites. The only storm that generated more than Irma is a hurricane called Ivan in 2004.
What can we expect from the rest of the season?
For the rest of the season, for a while early on we thought maybe El Niño would come on, and El Niño is warmer water in the eastern and central tropical Pacific, and what that tends to do is increase the shear of the upper-level winds in the tropical Atlantic, and shear breaks apart the storms. We thought originally El Niño was coming on, but now it looks like we've actually almost gone to La Niña, which is colder than normal water in the tropical Pacific, and that tends to reduce the shear, and typically leads to more storm activity late into the season, so into October. The season ends November 30th, but typically most of your big storms are done by mid- to late-October. I think the U.S. has only been hit by two storms on record in November. We're not quite to La Niña yet, but the overall environment looks more La Niña-like than El Niño-like. Given the fact that the Atlantic is warm and the tropical Pacific is fairly cool, the Atlantic hurricane season tends to get more storms later. No guarantee, but I don't expect the season to be done.
How unusual is it to get this many storms of this strength right in a row?
Obviously August to September has been very active, but it's not as much of an outlier as people think it is, because this is the first active September we've had in a while. From 2013 to 2016, if you calculated the ACE in those four Septembers, it was the quietest four Septembers in a row since 1911 to 1914. The U.S. hasn't been hit by a major hurricane for 11 years, from 2006 to 2016, and then this year we already had two. We went from running a streak of really good luck to really bad luck.
What role do you think climate change is playing in this bad luck streak?
One of the challenges is the observations of hurricanes are not stationary with time, so over time we're observing storms better than we used to. Before the '60s we didn't have satellites, so anything in the eastern Atlantic—unless a ship happened to encounter it—we didn't know it was there. If you just plot the number of storms, there's this massive increasing trend. But if you look at the storms that last, say, longer than two days, there's no trend, and that's just because we're observing a lot more little, weak storms than we used to. If you start looking at this ACE index that I look at, this integrated measure, and you go back to the late 19th century, there's no trend for the Atlantic. I think the problem is this: what we're trying to look for is that storms are going to get a little stronger in the future. We're talking here just about the wind. And you're trying to detect a signal that's not massive—we don't expect storms to be blowing at 400 mph, we're talking storms that are maybe 10 percent stronger. And if you're looking for a storm that's 10 percent stronger, trying to detect that kind of a signal in a time series that's so noisy, you're going to have a hard time doing it.
Its easier with things like global temperature: You plot it, and, yeah, there's a wiggle, but it's going up. There's not a lot of debate on that. Or with sea ice. There's some variability but its generally going down. But with hurricanes, especially in the Atlantic, there's these really large kind of multi-decadal swings.
What you can say with 100 percent certainty is storms will be more damaging, just given that there's more people and more stuff to be harmed. Obviously with sea level rise, you're going to see more damage, just given the fact that the ocean is six inches to a foot higher. Six inches or a foot doesn't seem like much but that can make a big difference in how far inland stuff gets inundated. You also have the fact that, in general, it looks like precipitation rates in these storms have increased a bit, so you have slightly heavier rainfall, which would also exacerbate flooding issues. And then obviously you have land use changes over time. A lot of places where once you had permeable surfaces like grassland and trees, we now have concrete.
With these things, you're going to see more damage and it's going to look worse even without climate change-induced increases to storm intensity, just through all these other human-induced causes like land-use changes.
This interview has been edited for length and clarity.