Pop Science: Hurricanes
So about a week and a day ago, I took some time out from the thankless writing of my midterm exam, to engage in a spot of scientific tourism. The Simons Foundation was hosting a talk on hurricanes and climate by Kerry Emanuel from MIT, so off I went. They have awesome tea, by the way (fruits! cookies! brownies! Teas! An Espresso machine!), and in a sinfully luxurious way, we were told not to throw out our own plates - there was a personage to do that for us! Needless to say, these are my impressions of the talk a week after the event, not any sort of faithful transcription.
There was an interesting intro (for me, who never really bothered with such things before) on the history of tropical cyclones. Evidently, they differ from regular storms since they draw their energy from warm water (regular storms draw their energy from temperature/pressure gradients in the atmosphere), and ye olde Coriolis force determines the direction of their spin, to form that nice characteristic eyewall. We found this out when some guy was riding around in the aftermath and noticed which direction the trees had fallen - early North American contributions to Science!
The Galveston disaster was regarded as impossible by meteorologists of the era because they thought that when the water becomes shallower, waves can't possibly go higher, or some such belief. Given how tsunamis and water waves work, that was, um, quite probably one of the more catastrophic scientific misconceptions in American history.
Next, we started getting down to the meat of the talk: how often should we expect catastrophic hurricanes in NY and points further north? Results from coupling current models for predictions of hurricane tracks to models of ocean currents and tides, and then running Monte Carlo while "seeding" miniature hurricanes, were presented, and basically, it's on the order of once several hundred years (b/w 300 and 700 IIRC, depending on how you count) Several caveats on this particular model: he was modelling only hurricanes, not storms that draw their strength from other sources. Sandy was a hybrid, and did both, and thus not accounted for by his model.
It was remarkable that, both historically and in the computer model, tropical cyclones are mostly a phenomenon of the north-central Atlantic and the north-west Pacific, with some much milder level of activity in the Indian and the north-east Pacific. The southeast Pacific and the south Atlantic were remarkable for basically having no activity (I think one storm in one hundred years, or something similarly low). A couple of people clued in to this and asked why - and since Jim Simons was one of them, followup questions were asked until the details were fleshed out - and basically we don't know. One of the answers offered was that given a different set of initial conditions but the same forcing conditions, maybe we could have a different climate model - which is another way of saying we really don't know. However, we speculate that it has something to do with that the northern seas are a lot saltier than the southern seas, that there is a net transfer of heat north, which cools the southern seas, etc. In other words, for beachfront property that has no danger of hurricanes, go to South America, or southwest Africa :)
Next there were some simulated results based on adjusting the parameters in his model to match with certain models for global warming over time, and a corresponding uptick in hurricane activity. Aside from the "expect more" statement, he didn't go into solid numbers, but I did some napkin calculations from his plots, and, well, in 100 years we might get a doubling of the hurricane rate, which, for once every 500 year storms, is not really terrifying. I really came away thinking that the putative influence of climate change on hurricanes is rather over-hyped, even conditioned on their models being correct.
More interesting was the discussion about the effects of hurricanes on the climate. There is some geological evidence that during earlier, warmer eras, we had much milder temperature distributions - warm at the poles, to be sure, but not ridiculously hot at the equator - that don't match with our current models of how climate should work. He offered up increasingly frequent hurricanes as a possible "smoother" of the temperature-by-latitude distribution. Hurricanes churn up a lot of water from the cold depths, mixing layers of ocean that normally do not exchange that much heat, and exposing the deep water to the light of day. They also throw a lot of stuff into the atmosphere where it freezes and then comes back down as ice - one reason why hurricanes look so fiercely white from space.
One other interesting fact: when hurricanes pass over land, they're cut off from their power supply, and one of the things this does is cause them to lift off the ground a bit. There evidently have been storms where at ground level (Central Park was mentioned), the winds were barely 20 mph, while they were dealing with hundred-plus mph winds at the top of the Empire State building, or at the top of a hill in Boston. Moral of the story: stay the heck away from skyscraper windows during a hurricane!
Anyhow, that was most of what I got out of the talk. It was fun, and a good distraction from the daily grind. Next talk is coming up in April, and I think I'll go again - this one's on the magnitude of disruptions caused by climate change.
There was an interesting intro (for me, who never really bothered with such things before) on the history of tropical cyclones. Evidently, they differ from regular storms since they draw their energy from warm water (regular storms draw their energy from temperature/pressure gradients in the atmosphere), and ye olde Coriolis force determines the direction of their spin, to form that nice characteristic eyewall. We found this out when some guy was riding around in the aftermath and noticed which direction the trees had fallen - early North American contributions to Science!
The Galveston disaster was regarded as impossible by meteorologists of the era because they thought that when the water becomes shallower, waves can't possibly go higher, or some such belief. Given how tsunamis and water waves work, that was, um, quite probably one of the more catastrophic scientific misconceptions in American history.
Next, we started getting down to the meat of the talk: how often should we expect catastrophic hurricanes in NY and points further north? Results from coupling current models for predictions of hurricane tracks to models of ocean currents and tides, and then running Monte Carlo while "seeding" miniature hurricanes, were presented, and basically, it's on the order of once several hundred years (b/w 300 and 700 IIRC, depending on how you count) Several caveats on this particular model: he was modelling only hurricanes, not storms that draw their strength from other sources. Sandy was a hybrid, and did both, and thus not accounted for by his model.
It was remarkable that, both historically and in the computer model, tropical cyclones are mostly a phenomenon of the north-central Atlantic and the north-west Pacific, with some much milder level of activity in the Indian and the north-east Pacific. The southeast Pacific and the south Atlantic were remarkable for basically having no activity (I think one storm in one hundred years, or something similarly low). A couple of people clued in to this and asked why - and since Jim Simons was one of them, followup questions were asked until the details were fleshed out - and basically we don't know. One of the answers offered was that given a different set of initial conditions but the same forcing conditions, maybe we could have a different climate model - which is another way of saying we really don't know. However, we speculate that it has something to do with that the northern seas are a lot saltier than the southern seas, that there is a net transfer of heat north, which cools the southern seas, etc. In other words, for beachfront property that has no danger of hurricanes, go to South America, or southwest Africa :)
Next there were some simulated results based on adjusting the parameters in his model to match with certain models for global warming over time, and a corresponding uptick in hurricane activity. Aside from the "expect more" statement, he didn't go into solid numbers, but I did some napkin calculations from his plots, and, well, in 100 years we might get a doubling of the hurricane rate, which, for once every 500 year storms, is not really terrifying. I really came away thinking that the putative influence of climate change on hurricanes is rather over-hyped, even conditioned on their models being correct.
More interesting was the discussion about the effects of hurricanes on the climate. There is some geological evidence that during earlier, warmer eras, we had much milder temperature distributions - warm at the poles, to be sure, but not ridiculously hot at the equator - that don't match with our current models of how climate should work. He offered up increasingly frequent hurricanes as a possible "smoother" of the temperature-by-latitude distribution. Hurricanes churn up a lot of water from the cold depths, mixing layers of ocean that normally do not exchange that much heat, and exposing the deep water to the light of day. They also throw a lot of stuff into the atmosphere where it freezes and then comes back down as ice - one reason why hurricanes look so fiercely white from space.
One other interesting fact: when hurricanes pass over land, they're cut off from their power supply, and one of the things this does is cause them to lift off the ground a bit. There evidently have been storms where at ground level (Central Park was mentioned), the winds were barely 20 mph, while they were dealing with hundred-plus mph winds at the top of the Empire State building, or at the top of a hill in Boston. Moral of the story: stay the heck away from skyscraper windows during a hurricane!
Anyhow, that was most of what I got out of the talk. It was fun, and a good distraction from the daily grind. Next talk is coming up in April, and I think I'll go again - this one's on the magnitude of disruptions caused by climate change.