Lessons from last summer's hurricane season
Last summer when Irene and Lee blew through New York, both were classified as tropical storms, not hurricanes. Yet they still managed to cause $1.5 billion in damage across the state. A year later, the cleanup and recovery is far from complete.
So how can these weaker storms still wreak such havoc? And did the classification of the events as tropical storms rather than hurricanes cause some people to let down their guard?
"In hindsight, there were several groups that were trying to communicate that the rainfall was going to be the big issue with this storm, but it didn’t always get into the public perception."
Dr. Ryan Torn is an expert in atmospheric prediction at the University at Albany. He talks about how scientists are getting better at understanding how extreme weather behaves, and why people don't always heed the warnings.
Note: this interview has been edited for length and clarity
Q: What is the most dangerous aspect of a hurricane?
A: It really depends on a number of factors, starting with where you live and the intensity of the hurricane. For a lot of weaker hurricanes, people often let their guard down. They think ‘Oh, the wind is weaker so it must not be that dangerous. I can stay in my house.’ Especially people who are inland.
In a lot of cases in a weaker hurricane, the bigger danger is the rainfall. There can be many inches of rain as we saw last year as we saw last year with Irene. In the stronger storms, if you’re on the coast, there’s a double danger: the wind speed and/or the storm surge.
There’s not a perfect relationship between wind speed and storm surge. You can have weaker storms that can create a huge storm surge, or you can have intense storms that don’t have a big storm surge.
Q: So a really weak storm that just sits over the land can do a lot more damage than a powerful storm that blows by the coast?
Q: How have weather modeling and hurricane forecasts improved? What are scientists still struggling to understand?
A: We’ve gotten a lot better at track over the past 20 years. We’re improving track forecasts on the order of a few percent per year. Over time, what used to be two-day track error, is now a three-day track error.
The leading reason we’ve gotten better is that we’ve been making models of the large-scale aspects of the atmosphere. We’re getting really good at taking observations of the oceans with satellites and incorporating those observations into our numerical models.
If we talk about the maximum wind-speed forecast of the hurricane (or intensity) we haven’t gotten really better over the last 20 years.
Hurricanes have these large-scale influences like how much water vapor is in the air, and the difference in wind speed at different levels of the atmosphere (or vertical shear). We’re pretty good at predicting those things. We also have a pretty good idea of what the sea surface temperature looks like.
What we don’t really know too much about are what we call “internal processes.” Hurricanes can go through a lot of natural fluctuations. There’s something called an "eyewall" replacement cycle where the hurricane can actually create a new "eyewall". It’s a really interesting process; the position of the eye (of the hurricane) changes.
In 2005, scientists observed this through aircraft really vividly in Katrina and Rita. We know how the process happens, but we don’t know how it starts yet, or what causes it to start.
Q: Another thing scientists don’t understand very well is how much rain a hurricane can produce— why is that?
A: One issue is data. Over the ocean there are no rain gauges for us to be measuring. Once a hurricane comes on shore, there are a lot of rainfall measurements. There have been a lot of studies, especially in the Southeastern United States, about some of the factors that give enhanced rainfall.
But it’s often easier to do hurricane research over the ocean. Land adds a big complication factor. So while you have lots of rain gauge data over land, it’s often hard to pick out how much of that rainfall is coming from a land influence versus the hurricane itself. Rainfall in general is a very hard problem for models. It’s one of the weaker links in numerical models.
In the case of Irene, there was a good sense that there was going to be a lot of rain. No one had a really good feel for what the exact number was going to be, but everyone knew that this was going to be a very large rain event three days out.
In hindsight, there were several groups that were trying to communicate that the rainfall was going to be the big issue with this storm, but it didn’t always get into the public perception.
Q: Is it difficult sometimes to get weather messages out to the public and make sure people act on the warnings?
A: There’s a whole group of social scientists that are really interested in this problem… giving people raw forecast data and then trying to see how they use that data.
With Irene, forecasters were putting out warnings a few days beforehand saying, ‘Hey, this looks like it’s going to be a lot of rainfall.’ The problem is I think sometimes the public perception is different in that the notion of a storm surge in New York City is a more exciting thing than rainfall over upstate New York.
So that just got a lot of media attention, but I would say the experts had a pretty good handle on this storm.
Q: Hurricane Irene was downgraded to a tropical storm before it hit New York. Do you think that caused some people to take the threat less seriously?
A: That’s absolutely a fear for forecasters. The category of a storm only describes its wind speed, it doesn’t describe all those other things. It doesn’t describe storm surge or rainfall. People can let their guard down.
This happened in 2008 with Ike in Texas. As it was approaching the Houston area, many people evacuated. Then the storm weakened. The problem was that when it had been stronger, it created a large storm surge over the ocean. Storm surge doesn’t respond to the instantaneous wind, it’s kind of a history of all that sustained wind over many days.
So all of a sudden, there’s this Category 2 storm, and people are thinking they can deal with that. But because it had been so large and so intense for several days before that, it came with a very large storm surge and caught a lot of people by surprise and did a lot of damage.
With Irene people said, ‘Oh, the wind speed is going down. Not a big deal.’ But again, they didn’t take rainfall into account. With weaker storms, the bigger impact is often rainfall, and rainfall will often have a bigger impact over a larger area than wind and storm surge will, which are primarily costal problems.
Q: Say we’d all listened carefully to the Irene forecasts. Other than evacuating, what can you really do about massive flooding?
A: In this kind of event, not much. In theory, we could have drawn the reservoirs down to a really low level, but you need some time to do that.
I know a couple of days beforehand, they did start opening up reservoirs, but this is just a really extreme event. When you get that much rain over that large of an area, over that amount of time, there’s just not much you can do, except to warn people.
Q: What role does climate change play in hurricanes?
A: That’s a very hotly debated topic right now. There are lots of things that go into intensity change in hurricanes.
One of the things people readily see is that the intensity of a hurricane is a function of the sea surface temperature. The warmer the sea surface temperature is, typically you’ll get a more intense hurricane.
All other things being equal, with climate change you might expect the sea surface temperatures to get warmer, and you might expect there to be more intense hurricanes.
There was data about five years ago, and several groups published studies showing that the global intensity of hurricanes has increased over past 20 years. They looked at sea surface temperatures and they saw they’d increased at the same time. So they said, ‘Oh there must be a relationship between the two.’ They predicted that hurricane intensity would keep going up.
It turns out that in 2010 and 2011 we’ve observed some of the lowest intensity hurricanes over the past 40 years.
That kind of threw everyone a curveball. There’s no good explanation as to why all of a sudden we’ve gone to a minimum, in terms of intensity. You have to start thinking about some of these other factors—like a difference in wind speed (or wind shear). The more wind shear you have, the harder it is for a hurricane to develop for a variety of physical reasons.
One of the things we don’t really know about going into the future, is how is that distribution of wind shear is going to change. Even if the sea surface temperature warms a lot, if the wind shear changes, we may not get any increase in intensity. It may get harder to make a hurricane, in general.
The scientific consensus seems to be that we should get more intense hurricanes in the future, but we might get fewer of them.
But I would say this is one of the great open questions in the field right now.
Mapping the threat
Researchers at the National Center for Atmospheric Research developed this map, which shows how vulnerable different areas are to hurricanes. They cite a study which shows that more than half of hurricane-related deaths happen in inland counties: