I think that paper is really important because during the 1970s, 1980s, and early 1990s, we were in a very quiet hurricane period. This was the same time that populations were growing very quickly along the U.S. coastal areas, as well as in Central America and the Caribbean. In 1995, we switched over to a very busy hurricane season. 1996 was busy; then it slipped back to quiet in ‘97, and then there were a couple more busy years. Partly it was that the number of storms increased, but major hurricanes—categories 3, 4, and 5—essentially doubled.

“Actually the amount of hurricane damage in the 1990s was about average, compared with the rest of the 20th century record, if you normalize for societal changes.”

In that 2001 paper, we were trying to take a quantitative look at how hurricanes changed, and relating that to patterns of ocean temperature variations, as well as changes in wind shear in the atmosphere, how winds change with height. Our main conclusion was that a), we had, in fact, gone back to a busy period in the Atlantic, Gulf of Mexico, and the Caribbean, and b), it was caused by a natural fluctuation in the Atlantic Ocean and the atmosphere, called the Atlantic multi-decadal oscillation. And we could say this because we had historical records for hurricanes and ocean temperatures, as well as other studies of proxy records of temperatures, suggesting that these busy and quiet periods tend to last 25 to 40 years each. Thus we concluded that it was likely to stay busy for another decade or two to come.

In that paper, we discussed hurricane records for the entire Atlantic going back to 1944. That’s the year that aircraft reconnaissance began flying into storms. It’s a fairly complete record, although it likely few missed a few in the 1940s, 1950s, and early 1960s, before satellites came along. We also went back to the U.S. record for land-falling storms, which dates back to 1900. There we think we have a fairly complete record of how many hit us and roughly how strong they were.

Ocean records went back a bit further, to 1870, so we had 130 years of ocean records to work with. And we had referenced several other papers by researchers who had done paleo-oceanographic work. These looked at core sample in the ocean going back several hundred years, and so they analyzed these cycles for quite a long time. Even though we didn’t have hurricane records before 1900, we did have indirect evidence of cycles through these ocean temperature changes going back several hundred years.

In the last year, there have been a couple of very highly cited papers that have linked changes in warming occurrences in the last couple of decades with hurricane activity both in the Atlantic and worldwide. These analyses suggest that there’s been a big jump in hurricane activity, due to manmade causes. So that’s certainly the big question.

In the Atlantic, for instance, where we have these very pronounced multi-decadal swings back and forth, how do you determine what may be part of a long-term trend, which may very well be due to global warming, and what’s due to these natural swings going on? That can be difficult to do, when you only have a hurricane record of 50 to 80 years total and changes that may be manmade on a scale of 100 years. What might look like a trend in the short record may indeed be part of a longer-term natural cycle. It’s a difficult determination to try to figure out what’s manmade and what’s natural.

That was work Roger Pielke, Jr., and I did. We looked at the damage that hurricanes directly caused in the U.S. going back about 70 years and tried to determine what’s going on here. In 1992, we had hurricane Andrew, which did $26 billion in damage. That was almost triple the previous record, which was held by Hurricane Hugo in 1989 with $8 billion. That’s a big jump. The speculation is that this jump in damages was perhaps manmade, again due to global warming changes.

What Roger and I tried to do is take a close look at the damage record itself. Maybe there were non-climate-related factors in that damage record. We tried to normalize the database, by asking the question, if hurricanes of the past hit today, roughly how much damage would they do? We took the original damage estimates, which are maintained by NOAA, and we said let’s first take into account inflation. The dollar today is worth less than it was previously. A second factor is per capita wealth. In general, people are wealthier in U.S. society than they had been in the past. So we got the Department of Commerce wealth index and, in general, we found that we all have twice as much as stuff as our parents did and four times as much as our grandparents. When a hurricane hits, both inflation and wealth cause more damage to show up in the record.

But the biggest factor, causing the huge increases in damages is the increase in population. The coastal counties, especially from Maine to Florida, have been rapidly increasing in the numbers of folks living there. Some areas have experienced a doubling in population every 20 to 30 years.

After you take into account those three factors, you find that the 1990s were not an outlier. Actually the amount of hurricane damage in the 1990s was about average, compared with the rest of the 20^th century record, if you normalize for societal changes. The 1970s and 1980s were actually the quietest era for damage. The big conclusions in that paper were that the 1920s, 1940s, and 1960s had more damage than the 1990s. And so once you’ve normalized that record, it starts to tell us something about climate. The implication you can pull out from that is that, a), there is no long-term uptrend in damages, after you adjust for societal changes; and b), once again, these very pronounced decade-to-decade swings show up in damages, too. It starts to give us a little bit of a clue about what the global warming impact on hurricanes may be. From that record, it suggests that there is no trend that can be linked to ocean temperature trends and global warming.

I’m not sure "combating" is the right word. For what we’re doing, "disentangling" is a better word to use. Society changes every year. Hurricanes that would have hit 50 years ago on any stretch of the U.S. coast would have a lot more impact today. It’s very difficult to pull out the real climate signals from this kind of impact data. I think Roger and I were trying to disentangle the reality from the historical records and I think we were able to do that successfully.

Two very prominent reports have been published recently: one by Kerry Emanuel in Nature ("Increasing destructiveness of tropical cyclones over the past 30 years," Nature 436 [7051]: 686-688, 4 August 2005), and the second by Peter Webster in Science (Webster PJ, et al., "Changes in tropical cyclone number, duration, and intensity in a warming environment," Science 309 [5742]: 1844-1846, 16 September 2005). The Emanuel paper showed a doubling of activity for west Pacific tycoons and Atlantic hurricanes when you look at an index that takes into account how many storms, how strong they got, and how long they lasted. He linked this big jump in activity to an ocean temperature rise that may be due to global warming. Webster’s paper looked globally since 1970, and it likewise concluded that there was a doubling of category 4 and 5 hurricanes around the world between the 1970s and the last 15 years.

Well, the Atlantic hurricane database is the best one, quality-wise, around the world. But even that has difficulties when you’re going back to the 1940s like Emanuel’s study did. Back in the 1940s, they had aircraft reconnaissance, but they only did storms near land. And they didn’t fly into the eyes of the storms once they got strong. Then the satellite data began in the 1960s, and everything changes.

That is the main issue in Webster’s paper. That’s looking globally, and they see this big jump in category 4s and 5s around the world. But what’s utilized is satellite pictures to make estimates of the winds. But the satellites have changed dramatically since the 1960s and 1970s, and so have the techniques to interpret satellite pictures. The concern globally is that there may not be a suitably reliable database to say what’s going on with these category 4s and 5s.

Moreover, a paper just appeared in Geophysical Research Letters, by Phil Klotzbach ("Trends in global tropical cyclone activity over the past twenty years (1986-2005)," Geophys. Res. Lett. 33[10]: Art. No. L10805, 20 May 2006) where he just looked at the last 15 to 20 years, when the global databases should be more reliable, and what he found was that the Atlantic, yes indeed, had a big jump up. But the Northeast Pacific had a decrease by about the same amount. And the rest of the world had no change in the hurricane activity. Yet ocean temperatures have gone up by about .2 degrees centigrade in the last 20 years. So these big global increases may just be an artifact of unreliable data; that we’ve done a better job recently in identifying the category 4s and 5s out there. That was very difficult to do in the 1970s.

To me, some of the researchers utilizing these databases are a bit naïve about the quality of this data and how it’s been constructed. It wasn’t put together by researchers, but by forecasters a week or two after the storm. There’s no quality control, or taking into account how databases changed over time because of increased understanding and new tools. Emanuel tried to do that, but the way he has gone after it is problematic.

  What would you like to convey to the general public about your work?

Well, maybe one message is that this issue about climate change and hurricanes is a real important one, and that there’s been a lot of controversy in our field, in published reports and in conferences about what’s going on. It’s certainly confusing to us in the field and it has to be incredibly confusing to the public at large. That really is a sign of an active area of science, and in this case it’s on a very important topic. I think this confusion is just natural. I think it’s a healthy thing. It’s a small field, too. The number of people actually doing hurricane climate research around the world might only be 30 or 40 total. There’s not a huge number of folks.

They are. More really smart people are starting to take a close look at both computer model simulations of today’s climate and future climate, as well as taking a closer look at the observational databases. People who were sort of interested but on the edge of the science are now jumping in and getting involved.

Yes. Yes. I worry about them a lot. I have a family here. During last year’s hurricane season, we got hit twice directly. Hurricane Katrina went over us as a category 1 storm, and then we got hit when Hurricane Wilma came ashore. We didn’t have structural damage in the house, but we lost some landscaping. And we had no power for five days with Katrina and eleven days with Wilma. We have good shutters on the house and we had a lot of supplies to ride out not only the storm but also the aftermath. So, yes, these hurricanes impact me and my family directly and so this is another reason why I want to know what to expect, not just next year but for the next couple of decades.

Christopher W. Landsea, Ph.D.
National Oceanographic & Atmospheric Administration
Miami, FL, USA