>> don, do you still usempc ? [inaudible conversation] >> c-span: where history unfolds daily. in 1979, c-span was created as a public service by america's cable television companies and is brought to you today by your cable or satellite providers . people in texas are preparing for hurricane harvey to hit. next, all the buyer for james elsner says ocean warmth is making hurricanes stronger and tornadoes are coming in with a more powerful punch. he spoke at the florida state university coastal and marine lab in mid july. >> thanks a lot felicia, that was, actually i wear more

than one hat also. i think of myself as a scientist, really. but if you do science for a long enough time , it puts you in positions like being the chair of a department itself. it's not something you seek out but it's something that happens because you do a lot of science so i like to think of myself as a person who just focuses on one thing. actually, it's two things. it's hurricanes and tornadoes so i focus on two things but they are really connected in that they are these violent windstorms that cause lots of damage and casualties and so we have to to think about what might happen in the future. so really, what i do and this is this idea of wearing one hat.i spend all of my time thinking about what

hurricanes and what tornadoes might be like in the future. so you're going to see a lot of science today in the sense that you're going to see a lot of grass. we use a lot of grass because that's the way we make comparisons, science is about comparing this with that. you're going to see a lot of grass, i'm not going to apologize because i think that is the window to understanding science but i want you to know that these are my grass. these are graphs that i got off the internet. these are someone else's arguments about what's happened. and there are a lot of arguments and there are a lot of opinions. what you're going to hear from me is what i do and what i think about on a daily basis. as felicia said, i'd like to before i begin thankfully, thank felicia for inviting me to do this. i got a chance to tour the facility this afternoon, wonderful facility. this is really outstanding

facility that fsu should be very proud of and folks associated with fsu should know more about this fsu coastal and marine lab and a large part of the reason for that is because of what felicia has been able to do over the last decade so as she said, i share the department of geography, it's not a big department. we got 10, 12 faculty depending how you count but it's a very dynamic and increasingly associated with what to do down here at the marine lab. so i'm going to start out as i said two things, i'm going to start with hurricanes though if you dripped off in this and you might, when i get to tornadoes you know i'm about halfway done. so let's talk about hurricanes to start off. you are obviously familiar with hurricanes in a generic

sense. we can track them, we can look at them from space and they are these powerful things so we know what hurricanes are like today. last year we had atornado , very close to this part of the world and so they are in our consciousness constantly. when, where, how often? these are things we know about if we live near the coast but what about the future? are we in for a greater risk of these storms? unfortunately there are no simple ways to get answers to these questions. they are very important, there's just no simple way. why can't we just do some kind of calculation on the back of the envelope and work it out? first of all, the theory is very limited. don't have a theory of climate. this is how climate works, it's nonsense. we have parts of theories of

how things work but we don't have a general theory of how climate works. we certainly don't know everything about what drives hurricanes, asic stop so theory is limited, that's why it's not easy to do this. laterals which are good at forecasting where a hurricane might go given that there's one out there, they don't really represent the atmosphere and the ocean in an attic adequate way, at least on the scale of climate so finally we don't have enough data. we don't have a way to look back in the prehistoric times. it's been difficult to look back and we just don't have enough data. and the data that we do have varies in quality. of course social media is not going to help us out. you're going to get a lot of opinions and bickering to what is the solution? how does james elsner spend his days thinking about this problem? i tried to put things

together. i put that. together with the models in the data so i try to combine. i'm going to talk about two theories today, thermodynamics and i'm going to talk about distance. the two theories that allow us to get some answers about what hurricanes might be like in the future. okay?so you probably aren't a statistician so i'm not going to go that the but these are the deep structures in which we can hang our hats on to try to understand what it might be like in the future. so we will start with thermodynamics. this is my thermodynamics slide from popular dynamics that basically describes how a hurricane operates. all you have to keep in mind is a couple prepositions. the first one is in, the

second one is up and the third one is out. in, up and out. that is the circulation of a hurricane that you are probably not aware of because you think of circulation like this. you think of these things like spinning and that's the wind is going to destroy your home, produce the surge, flood your house but that isn't the circulation that thrives the hurricane. it's the in, up and out so if you lead with these three prepositions, i've done my job. in part is where the air comes in at low levels, near the ground. near the ocean's surface and it gets its heat and moisture and then it rises in the center of the hurricane and also in all the other thunderstorms rounding the hurricane and it exhausts that higher level. so it takes in heat at high temperatures and exhausts the heat at much lower temperatures.

so it's not even kind of, it's almost exactly the opposite of your refrigerator. that's how things stay cold in your refrigerator. here you are exhausting at low temperatures. this is what we call a heat engine. and in fact this is an extremely efficient heat engine. you see prepositions describe the hurricane as a heat engine. it's a way of converting the heat and moisture of that warm ocean into the winds that you feel circulated. in, off and out.so that is a theory and it's based on thermodynamics that was worked out in the 18th century and it is called the carnal heat engine so with that heat engine., with that theory we can work out how storm or hurricanes can get. i'm going to use this npi to abbreviate how strong can

those winds rotate. 60 miles per hour, 80 miles per hour, 150 miles per hour. that's the speed, and going to call it maximum potential intensity. and that's just related to how warm the oceanis. sst is an abbreviation we use for how warm that ocean is. the warmer the ocean , the higher the intensity of the storm. they are proportional. warm ocean, more energy. so this is basic. this is worked out by a fellow by the name of carrie mendel, a good friend of mine at mit. he worked out how hurricanes intensify and how they get storms and the maximum potential derived potentially from how warm the water is. when you go swimming in the ocean tomorrow, feel that heat. at the heat. because of this engine.

that derives these hurricanes. you say well, that's kind of simple. we can figure out how warm it's going to be and we can figure out how intense the storms are going to be. if you'rethinking ahead, you're thinking why is this so complicated ? the warmer the ocean, the more energy. first of all, you have this thing here, the denominator. at the upper level temperature. how cold it is, that's downstairs. how cold it is, the stronger the storm can be. warm that the service, strong storm. even if the ocean is in that warm you can still get a hurricane. you get arctic hurricanes, right? if this is cold enough, even if this is not that warm but still you'd say he's got everything bear, why does he need anything? he thought his erie here, carrie manuel is his friend, he can work this out. but here's the problem right here and i'm going to evaluate elf for boundary layer flexors. this thing is that the

hurricanes start to mix up the ocean. they start to produce sea spray so there's a lot of complexity going on. it's not clear how much this is going to go up for a given disc is what we want to know because we know the oceans are warming up. they are warming up primarily due to greenhouse gas. there warming up, we should get stronger storms but how much stronger? who knows, because of this. we can't work it out. we can't work it out because of the boundary layer flexors. so that's one theory. bring in the second beer he. first theory again just to review. in, up and out. this theory of how storms intensified through the heat engine. your car has an engine, it works on a very similar idea. not nearly as efficient as

a hurricane. the other theory is statistics. this might even be further from your experience but let me see if i can humor you on this. it's a statistical theory worked out in the mid-20th century, about 1955 or so. the pose for example we record the highest wind speed in 10 consecutive hurricanes. so there's a hurricane out there now, let's say and after two days, it's gone that we know how strong it got. that's say it got to 34 and a half meters per second, i'm going to use meters per second. why am i using meters per second? i'm a scientist, i apologize for that already but that's what we use. if you need to convert, you

just double that, that's miles per hour, about 60 miles per hour or 70 miles per hour, how fast you go in your car. this was the first hurricane, that's how strong it got. the second hurricane hit puerto rico and died, it had a maximum intensity of 44.2 so we can do this. for each hurricane out there, tell me how strong did it get ? you're just going to put all those strengths down here. some get stronger than others, right? this is the set of 10 wind speeds for the last 10 hurricanes. now, what we can do is order this. this is how they occurred in time. this is more interesting from a statistical point of view. this is the weakest, this is the strongest. just order them all. this tells us that 20 percent of the hurricanes of these last 10 have winds that exceeded 61 meters per second. these two in red with a 20 percent of the hurricanes that exceeded that amount.

of wind speed, right? we have 10 percent that exceeded this amount so we have percentages and we have wind speeds. those two things make statisticians drool because they can connect the dots. they can connect the dots into this extreme values. so we can work out these the radical highest wind speeds by looking at what we call the limiting intensity. this gives us a limiting intensity from the data so from kerry and manuel erie of thermodynamics we can work out the maximum potential intensity and fromthis ethics we can work out the limiting intensity so we have two different intensities . >>

i want to know that. i want to know what would be

the tallest, so what i'd do, i'd have another lecture next week and get another 25 people, a different group of people and another tall person, now i have two tall people and that's what i'm doing here. i'm getting the set of tall people and sitting a curve through that and extrapolate to get the possible tallest person. that's wonderful theory. it's embedded in the mathematics. i've got two things, a statistical quantity to use to compare with the theory of maximum potential intensity. hopefully, you're still with me. i am maximum potential intensesy from theory, statistical limiting intensity from statistics. how should we make that comparison? okay? it turns out that the limiting intensities, the absolute, the tallest person isn't that person, it's how, how limiting

intensity changes with the ocean temperature. that's the key component here. so, how do we get at this? well, it turns out, hurricanes, of course, occur over oceans where the temperatures aren't uniform. so, maybe if you pay attention to hurricanes you'll recognize this season. this is a plot of all the hurricanes in one season. i didn't label which season it wasn't 'cause i wanted to, as a professor i want to test my students. what season is this, if you remember? it was a pretty active season. close, very close. 2005. okay? 2004 is an excellent guess 'cause they look very similar. there is 2005. we had lots of hurricanes. we got hit a couple of times here in florida. so, but you notice underneath the tracks of each of the white

lines, actually a series of hourly dots of where the hurricane was, at least the center of it, underneath that, those tracks are the ocean temperatures, you can see they don't all occur over the same temperature, right? so, what we can do is then use space, and this is, as felicia said, i was trained as an atmospheric scientist, but i really became a scientist when i became a geographer. the reason is, although atmospheric science is an important discipline for understanding how the atmosphere works, geography allows you to put pieces together because it allows you to leverage space, okay? it allows you to think about things spacially. when you start leveraging space you can get a lot more bang for your buck. let me show you what i mean. here is-- we can grid up the-- we can grid up the domain. okay. this is our atlantic basin. we can recognize florida here.

this is where our hurricanes form. and so, we can look at, for example, two hurricanes. one here, and one here. and the hurricane hours are gray until it gets into one of those hexagons and then it's black. it's not hurricane until it gets to block and then hurricane intensity. it's a weak storm and then becomes a hurricane. so what i can do, i simply can count how many hurricanes occurred in each of the boxes, each of the hexagons. this is another key component of my talk, besides the three prepositions hopefully you don't forget. the frequency of storms and i'm going to count this here. these boxes had one, these boxes had two, everyone else had no hurricanes.

but i can do that for all the hurricanes that occurred over many years. this is a 50-year plot. you can see where, where hurricanes are most common in this box here, just off the north carolina coast and this box here and i bet, especially this box here, comes as somewhat of a surprise to you. this is where more hurricanes occur than here. more than here. more than here, more than here. there's more hurricaning out here and here than there are in here. so, it isn't about frequency, it's not just about frequency. frequency doesn't give you the answer you might think. but-- so now i can do this. this gets complicated and humor me for another five minutes here. i take the time period, 1981 through 2010, okay? en this is simply the number of hurricanes that occurred in these boxes over that period of

time. okay? the darker blue indicates more hurricanes, the lighter blue, fewer hurricanes. pretty simple. i'm going to point to two boxes, c, d. and now within each of the boxes, i can tell you how strong the hurricane was when it was in that box. this is like getting back to the theory, i'm counting how strong each hurricane was. so each gets to have one value when it goes to that hexaegon. this is where the storms are strong and frequent and so, fewer and fewer,more hurricane,

not as strong. that's the key to this puzzle, okay? okay? >> i know it doesn't sound like he's going to say, i don't work at that scale. what i work at is the scale of trying to understand how the climate conspires to create hurricanes in this aggregate. so, this is how i work it out. i have these two boxes, and i can say, for example, for box c where there's fewer of them, i get a stronger limiting intensity and for d, where there are fewer, but stronger-- i mean, yeah, for d where there are more, but less intense, i get a lower limiting intensity. this is only about 50 meters per second. this is over 70 meters per second. okay? so i'm halfway done. sorry about that, only halfway

done with the hurricane because i only have the limiting intensities now. i have the nice spacial geographics-- meteorologists would never think about this, but geographers would understand this, exploit space to understand things happening over time. i have my limiting intensity and put it for each box and i have it for the ocean temperature. a-ha! now i have the two pieces i need. the limiting intensity and ocean temperature and put them together and i have the beautiful plot. here is my ocean temperatures, here is how strong storms can get and that slope represents the sensitivity, the sensitivity of hurricane intensity to ocean warmth. and it turns out that that is exactly, or close to each meters per second, to translate that, that is about 16 miles

per hour for every degree celsius of warming or two agrees of warming. that's how much stronger hurricanes we can expect to get just based on combining the theory with the data, okay? that's a beautiful result because it's never been worked out before how to get that sensitivity and it is a fundamental component of how the atmosphere works. and produces. so, why is this important? why? so, you say, well, okay, that's eight meters per second, plus or minus one. why is that important? well, turns out that the strongest hurricanes are getting stronger and they're getting stronger at about one meter per second per decade. that doesn't sound like a lot, right? you can see the highest, strongest storms, so for every year i can group the storms by their intensity. this is the median intensity for each other.

you can see that that's not changing much, but the upper kwan tiles are going up and they're going up at about a meter per second per decades. and so, this is where it starts to get interesting because the losses increased by 5% by meter per second. here is your wind speed and economic damage. and those say how much damage we can expect per every increase in wind speed. this is important, folks, because it maps onto the next aggregation of the next 20, 30, 50, 100 years and more of hurricanes we're going to see this kind of change, okay? forget about the fact that we're going to maybe build more stuff, we're going to have more stuff in the way, this is completely independent. this is completely based on the theory of how hurricanes operate.

so it's extremely important. okay. so i'm about halfway, so, just a real quick summary. we can understand what hurricanes might be like in the future come beaning the theory with the data. the theory, in, up and out, how hurricanes are heat engines and the data, this wonderful extreme values theory. and it looks like it's likely the strongest hurricanes get stronger per degree of ocean warming. you can take that to the bank. and this is one meter per second per decade, 5% increase in losses per decade, independent of how much we are exposing. if we're exposing more, of course, you're going to get more of that. but keeping exposure constant. and you say, jim, you took in one factor, sheer upper level temperature, et cetera, if we

put those in our model and we've done this, they're not really strong, they don't reflect that eight years per second, and there could be something i'm missing and as a scientist you reserve the right to be wrong and to be found wrong. that's part of our job, but because it's part of our job, we think about this all the time. okay. let me-- let me talk about tornados and we can stop and ask me questions. tornados are somewhat mysterious to you compared to hurricanes, but, as a scientist, to me, it's the same idea. that is this idea that the atmosphere produces in these very episodic ways, very extreme events, whether it's the hurricane, or whether it's the tornado and the tornados, of course, are a much smaller phenomena, but the winds blow much faster on tornados on the average than hurricanes. and there's been a lot of talk

about what's going to happen with tornados, and generally, the focus is on just looking at the annual counts. if you look at the number of ef-1 plus tornados, those are, so we rate these tornados on this safer simpson scale-- or the fujita scale and when it starts at zero and ef-1 tornados win about 40 meters per second or 80 mile per hour. so 80 mile per hour hurricanes are stronger and we don't really see any trend in the numbers much storms. so, this absence of trends in tornado occurrence doesn't imply that the tornado climate is stationary, okay? but it often stifles the discussion. so, what we find is the number of days with tornados provide additional information about what, how a tornado might be occurring. unlike hurricanes, we don't

have any theory to hold-- hang our hats on with regard to how, what tornados might be like in the future. we do understand the data and there are a lot more tornados than hurricanes, so we have something to go on there. unfortunately, we have become much, much better at observing tornados than in the past. so if you look at a long record of tornados there's generally an increase in numbers, but that's because, we probably are able to communicate what we see and there's just more people paying attention. so, we have what's called the population bias. and this graph really clearly shows this population bias. this is distance to nearest city center ands this the number of tornado reports scaled boy an area of 10 square kilometers. how many tornados do you see within 10 square kilometers? as a function of distance from a city, every city, this is every city, and you see a

higher rate about, between 1.2 and 1.4 tornados per 10 square kilometers and near the city to less than one about .8 away from the city. now, if you want to go out and say cities cause tornados, you can get front page headlines, right? but clearly, you probably want to step back and say, is that the causal mechanism? maybe i've got this thing backwards, right? of course, it's likely that this is due to the fact that people say, well, the cities, the cities are where people can report, it's an efficient mechanism of getting observations into record books, and so, what's what we're seeing here, but what's really interesting, i did the same thing in 10-year periods starting with '61 through '70 and going through the decades one year at a time. one year at a time in order.

what do you see in the plot as you get down, what do you see there? a couple of things, very interesting, okay? it me, this is-- i call this a snake plot. they look like these snakes, right? snakes that have been alerted to you. and so, what's happening to those snake plots over time? maybe it's hard to see this? it's flattening out, right? it's starting to flatten out. we don't see much urban effect relative to a rural effect today than we did back here. and it's higher, and it's higher. these are the two amend points of that plot, beautiful. but clearly, it isn't the cities that are causing the hurricanes because this would not go flat if that was the case. right? so, i just bring this up, folks, not because i like to look at snake plots, but i do, i do spend a lot of time making nice plots, okay?

but because that's how i think. that's the way i try to put the pieces together. i start to say, well, if i did it this way, i should see that. if i don't see that, then i'm surprised. okay. so, here is your number of tornados over time, since about 1954. and this is only the stronger tornados. and you can see it bounces around from year to year. sometimes we've had as many as 900. 2011 was a big year. but there's a lot -- there's no real trend in that, okay? there's no real trend. you couldn't say we're seeing more tornados, the climate is warming up. clearly the climate has warmed over this period by at least a half a degree, but we're not seeing more tornados, we're not seeing fewer tornados. we're seeing fewer tornado days. we're actually seeing fewer tornados-- the same number of tornados, but occurring on fewer days. this is work that i did with my wife, and she's sitting in the audience. she's very instrumental in

trying to get me to make these plots. anyway, we're seeing fewer days with tornados and that's the key, folks, okay? so, suppose i count the number of days, the number of days with at least tornados, i told you i'd have lots of graphs. the number of days with at least n-tornados. n is here, there are at least four tornados. there was about 35 days in 1954 where there was at least tornados reported at least four tornados i should say. right? so that bounces around, but it's -- as we get to larger n, double n to 8 and then to 16 and then to 32, you start to see this upward trend, okay? so what are we seeing, folks? we're seeing although the number of tornados is flat, the days in way there are big outbreaks is increasing.

okay? so, it's like the atmosphere saves itself for big day. to put it in and tlhrow-- anthrogenic terms. we can talk about the probability of the day with at least n tornados, n being four here, it's going up. eight it's going up, n equals 16 is going up more and n equals 32 is going up, way up. okay? so we're just getting these bigger outbreaks. so, what's happening? that's where my head is when i wake up in the morning, so it's really thinking about-- why is this happening? folks? well, that is the -- that is the, so i can do this a slightly different way, i can think about it in terms of percent change, in these different decades and again, no matter which way you do it, it

isn't a number of the events, but it's the big events. it's kind of like the hurricane problems, the strongest storms are getting stronger, no doubt about that, i quantified it. we're not seeing more hurricanes, but the stronger ones are getting stronger. we're not seeing more tornados, but they're coming in bigger bunches. that means we have fewer days with tornados, but when they come, they come in bigger bunches. all of these graphs are showing that same kind of thing. there's a large scale hypothesis put out. as soon as i put out these results, i had a lot of folks tweeting back and saying, simply, well-- i think about this as efficienciment i actually like that way of thinking about it. i like that way about thinking about hurricanes, too. and this year, it's more efficient. when you think about the stronger storms getting stronger, it's about efficiency, folks, it's about taking the time--

the time in which hurricanes can intensify isn't going to change. why isn't it going to change? because they have to come off africa, they have to move across the ocean and only have so much time to get strong. that's not going to change. the oceans aren't getting wider, at least on the scale of humans, so, so they're going to have the same amount of time and they're getting stronger, which means they're more efficient. so i like this idea of efficiency and that's the same thing, what's going on with tornados. we're not seeing more tornados, but they're-- when we do, the atmosphere tends to be more efficient at dropping them from the clouds. so, we called this the large scale pop. because there's a large scale dynamics, like wind sheer might be causing. so my first thought, based on this large scale hypothesis, is that the area, so we've got a big day. so, the area over which the tornados are dropping out of the sky, they don't actually drop out of the sky, okay?

tornados actually start from the ground up. okay? pro tip. they start from the ground up. most tornados, not all of them, but most tornado. i think the water spout is probably different, but the tornados that we're talking about start from the ground up. they start to spin. it looks like they come from the cloud because tcondensation comes from the cloud. that's what we see first. the spin actually starts at the ground. it's spending up here, but on a much larger scale and then it kind of gets together and then speeds up. anyway, we can have another lecture on exactly how tornados form, but here i'm talking about, i'm thinking that maybe the atmosphere is just getting better over a larger area. so, let's look at a particular dayment this is just one day in 2008, may 8th, and there were a number of tornados, but they'd occurred in two different regions. so, what i do, is he think about this as a cluster of

tornados, and this is a cluster of tornados. i just draw a box and again, i'm a geographer so i make these maps and draw out where the tornado occurred. and what i think is happening, my hypothesis was, these things are getting bigger, there's more of them so these areas are getting largers. the area of the clusters is getting larger. this is an example of clusters. so i put this red dot. this is the tornado that represents the center of the set of tornados. it's the closest to the center. this is the closest to the center. so we call that the middle tornado. we know about the median or the mean. it's like the median, it's closer to the median. it's the tornado that represents the closest to the center. so, i'm going to get rid of all tornados. whoops. yeah. [laughter] >> okay.

okay. okay. >> okay. well, anyway, it's not so important. what's important is that i was wrong about my hypothesis, that is, this is the mean number of clusters per tornado day, so we're not seeing more clusters, that's pretty flat. although it's going up, it's not really significant. maybe there's a few more clusters, but it's not. and the total area of clusters is flat. so, i was wrong about this large scale hypothesis. so, what's happening, folks, is we're getting more tornados. we saw that, we're getting more tornados, but they're not getting bigger area, they're just more of them occurring within a single area that's favorable for formation. and so this is-- so then i can look at the density of-- how many tornados per cluster, tornados per 10,000 square

kilometers for four, eight and so on days and you can see that that's a tremendously significant upward trend in the efficiency of tornados. okay? so this is a little disconcerting. and i know you're saying, well, why do you think that's-- i don't know the answer. but i think it's-- we hypothesize, driven by cape and-- it's not a large scale. it's probably the local scale. et cetera probably a local scale phenomena making the clouds more efficient at producing this rotation. so, i think it's back around to thermo dynamics. that's what i like about it. all of thermodynamics is local. when you think about the moisture is warmer than it was years ago, it's more humid

therefore we can get more efficient tornados as long as we have some kind of cap which keeps tornados from forming on more days, but when it does pop, you get lots of tornados. again, speculation at this moment, i don't have the answer. see me in about a half year, i probably will. okay. so i've gone on for about 45 minutes. i'm going to say one more thing, are tornados getting stronger? we didn't talk about the strength of the tornado, we simply talked about the numbers per area and numbers on the big days. the problem with the intensity of tornados is we don't really measure how fast these tornados are rotating. they just-- fr first of all, they're rotating so fast, if you had a wind instrument like an anamometer, it would probably be destroyed. there are a lot of chasers that

troy to put their cars and instruments in the path of tornados, but those are very few. you're not going to get too many intercepts like that. so, what we did, we thought, well, we've got to be able to back out the intensity of the storm from this damage scale. and so that's what-- so you can count the number of tornados, but it doesn't tell you how many. you can count the number of ef 3's and ef-4's and it doesn't tell you if the strongest ones are getting stronger. here is an example of a tornado path. that is what we call the damage path. here is oklahoma, oklahoma city here, this is the moore tornado from a few years ago, that was the damage path in gray. okay? so, there's a couple of things you can see about that path. first of all, it's-- we can think about the area of that, how much area does that tornado cover? we can characterize the area by

how long it is, and how wide it is. so length and width turn out to be two ways to describe the path of a tornado. so we look at the different tornados, these are stronger and stronger from weak to strong tornados. this is the number of tornados by that category. there's a lot more weak tornados than strong tornados. only 14 ef-5's since 1954, but here is the area in square kilometers. okay? so we can see that the big tornados are much longer. that makes sense, right? a tornado doesn't come out of the sky as an ef-5, it comes out fairly small, rotating maybe 30, 40 meters per second. it can get up to 120, 130 meters per second, but it takes some time. so, it travels over longer distances, so the path area is much longer.

so, if i know a tap area, you can work out an energy of that storm and so, this is the tornado energy, okay, over time. so, here is our year, startening 1994, through 2016. and we can see the approximate energy of the tornado ef-1's has gone way up. so tornados not only are they coming in bigger bunches, but they're expect stronger, longer wider and thus more energetic. okay? which is really, really interesting, but nobody's got any answers yet on that. so, if i did this just for florida, we're seeing longer paths in florida. okay. it's like to play with plots as you can see. so i decided to do it a different way. here is your year going down. this is late to early, to today, okay? and so, this is the distribution of path links. this is how long they are, and

you can see that over time they're getting longer. you can see the hump is moving to the right. that's actually on a long scale. so that's a fairly significant change, not uniformly overtime, but you can see that they're getting longer. these are just this part. okay. so, some final thoughts. single metrics can be misleading when it comes to climate change. when people say we haven't seen a hurricane forever, that means climate change, no, not really. they say, we've-- we used to see just as many as before or we see more now, that's probably not it. that's not it. and these are single metrics. how often and how strong are really the two conical components of storminess. take any storm. think how frequently they occur and how strong they are when they occur and by analyzing

them together, we can get this kind of broad understanding of a storm climatology, whether it be hurricanes or whether it be tornados. so, with hurricanes, typhoons and specific tornados in the u.s., the fingerprint of climate change appears to be fewer, fewer, but stronger. >> and so, this is a take-home kind of graph to try to understand what global warming might be doing to hurricanes, okay? and maybe storminess in general, right? so we're kind of thinking about activity going on its merry way, some years active, some years inactive. there's el nino events and tornados and climate variations going on and we may actually see falling occurrence of events. fewer days with tornados, fewer with hurricanes, but when we get them they become--

they're strong. the separation between frequency and intensity is really, maybe, maybe what climate change is doing to storminess. okay? and i'll leave it at that and open up the questions. [applaus [applause] >> [inaudible] >> in the blue. >> what you just said about the tornado path. >> yeah. >> was the f-scale, rating of

that order was? >> that's a great question. the 1925, and that is actually that occurred before the record in 1950's. but, yes, it is across several states, illinois into indiana, yeah. the ef-scale was not intented there. the f-scale was invented by theodore fujita university of chicago in the 1970's and it wasn't implemented by the national weather service until the '80s or late '70s. i think what they're able to do with tornados like that is look at photographic evidence and kind of say based on those photographs, the targets that

were hit indicated ef-2, ef-3, or ef-4 damage. i believe it's probably in the record book as an ef-5, but i'm not positive about that. that's a great question. >> the microphones are against each other. >> you guys don't believe me. >> that doesn't mean you can't have a question, but hold the microphone. >> i'm wondering, have you really looked at the data collection of all of the information because collecting more data every year and go back to like the communications data collection and all of that

in some ways compared to now. have you looked at how that -- possibility? >> that's a great question. and it is a question that i think about quite a bit because it is the-- you know, it is the question that if i'm wrong about the data, if i'm mistaken and if i'm assuming that these are this way and they're not, they're some other way, then, of course, my conclusions are going to come tumbling down. one of the things we do in my lab, is we go out and try to survey the tornado so we look at what they do in the modern time currently. so that's one thing. the other thing is when we build the models, we take into account things like the ef scale was implemented this year, so we can-- as a stas statistics, i can get

all kinds of stuff that could be there, but a lot of the stuff tends to kind of be random variation. so, it's not systemic. if there's any systematic, my models will catch it. that's the thing about models, they will catch the systematic and leave the variation as a residual. but i could interpret-- i can interpret nonrandom, i can interpret signal in the wrong way. but i try to work with folks that know a lot about the data set and i'm aware of a lot of the changes that have occurred over time, and so-- but those are legitimate

concerns. >> yeah, okay. so people had talked to me before, the folks at storm prediction center said, you know, these data have a population bias. and i said have you ever quantified the population bias. by doing the snake plot i can quantify it, for any region i can tell you when it's unlikely to have any population bias, at least population bias in the sense of rural to urban gradient. i can do distance near roadway. i put that roadway network in there and i say, okay, does the roadw roadway give me a better coverage. i can play the what-if games with the information i have. that's why, i think getting back to this idea about theory and data, if you have theory,

you can start doing things a lot. if you're just playing with dates at that, those are the caveats, buyer beware. >> go ahead. >> two quick questions. globally, who are your partners in researching the tornados and hurricanes and then, just a second to that, in this area, one of the things that the data has been used by the insurance agencies, is to sort of show the risk incrowing in the area in a lot of insurance companies that have left. so, do you see them taking your data and using it in a way that you didn't intend? >> that's a great question. it's really not simple to answer. there's a lot of things, first of all, about your-- about the people that i work with. as i mentioned, terry manuel, he's one of the guys i can lean

on. i would say he's at the proper distance from me. we're not friends, but we're good colleagues and we can-- but i pay attention to his work and he pays attention to mine. i've worked in the hurricane area with folks that have done some of the paleo work and looked at coastal works and marshes, and drilled down and looked for evidence of overwash deposits and pre historic events. you get the sand layers and deposits and sediment core that look at an event like a hurricane. and i tend to be fairly isolated. i'm kind of a-- i work in my-- my expertise in statistics tends to keep me somewhat isolated, but the tornado people, i don't-- i work a little bit with a guy

by the name of grady dixon, he kind of tried to redefine tornado alley to include the south. the south has a problem with tornados, not necessarily the coastal south, but the mid south has a problem with-- so that's a person. your second question about the insurance. a wonderful question. i have worked quite closely with insurance folks, not paid by them directly, and my work has not been funded directly by insurance companies, but they pay a lot of attention to what i do. i think, first of all, with the problem with tornados, they tend to underestimate the risk. tornados are a lot more common than people-- in other words, if you look at insurance policy in oklahoma you're probably-- well, i don't know about the policy, but i do know your risk is probably-- the risk of your house being hit by a tornado is about one in a thousand years.

that's what they calculate. i believe it's closer to one in 500 years, it's about half. i think they're underestimating risks of tornados in many places in the u.s. concerning the hurricane problem, and the coastal rates, i think the insurance companies has got it i don't think and i was part of that discussion. they came to us back in the mid 2000's and said what's going to-- 2004, 2005, big seasons, they said to me and a group of us and said, what's going to happen? we're seeing this. we said basically we're seeing a lot more storms in the atlantic, the atlantic is warming up. they took that idea that the atlantic is getting warmer and transferred it to the coast. we tried to stop them. we said, you can't-- the correlation between coastal activity and activity in atlantic is very weak. you can have big seasons without any hurricanes coming and that's the piece that they missed.

but of course, it was in their favor to raise the rates and they had the science to back them up, but the science on the coast was never there, that the storms are getting more frequent. what you do see though, if you look at losses along the coast and i've done this. if you look at historical losses, assured losses from texas to maine, the number of loss events is not going up over time, but given a loss event, the magnitude-- how much loss depends on ocean temperature. the warmer it is, the more losses you get. controlling for inflation and controlling for everything that we know of. so the stronger storms are produced by the warmer temperatures, that's not in doubt and it shows up in the loss events, but whether we'll see more loss events, i don't buy it. >> a general output on these

storms, and maybe [inaudible] and more generally, things that interest me-- [inaudible] . it's been around for a while and i was wondering if the theory could be somebody in that direction and looked at it-- >> that's a great. an excellent point and i think you're onto something, i really do. i know, i want to put this in kind of anthrogenic terms or organism, but the atmosphere has an ecological constraint. if there are going to be fewer of them there have got to be more or stronger.

and that occurs with things like rain. so, what is going on here? the system a being fueled and there's these counter forces that, you know, keep things in check, but when they go, they go hard. and again, all i can offer is a few vague terms and maybe borrow from ecology at the point. i think it's thinking that way to maybe unify the science of climate change and storminess and maybe it takes some biologist. maybe it will take some ecologist to help us out, think through this problem, but i absolutely love that idea. >> [inaudible question] >> are they going to get farther apart and bigger when they happen. >> right. >> is that in the-- >> well, i haven't studied that problem, but i think you could

do similar kind of diagnostics. i think the metrics -- i'd love to give another talk about that because i've actually done that, i've set up this space of frequency and intensity and mapped that onto climate change and it is this notion of efficiency that maps onto climate change. and it's exactly that and it would be fun to do it for a bunch of different things like drought and floods. >> because you talked about the weather. >> right. >> the climate, everything throughout the country, has that changed? seems to me we've seen a change in the past 20 years. >> yeah, yeah, i don't know the

answers to those questions. i do study them to the extent that they might impact the storminess that i think about. i mean, you've heard the term arctic arctic amplificatioamplificatiod of feeds into this idea that you will probably have fewer-- see, to get a tornado you need two things, you need wind sheer, which is kind of the opposite of a hurricane,right? a hurricane will get torn apart by wind shear, but the rotation is by the shearing winds. those shearing winds are caused primarily by the jetstream. so you're going to get a weaker jetstream. that would argue we're seeing fewer tornados. you're not seeing fewer tornados. but you need moisture, too, so

the moisture is going up, the jetstream on average are getting weaker. when it does come down, now you have more moisture. in a hand waving way you can explain my results, but i tried to connect them, but i was shot down, but that's what i do, so that's fine. yeah. >> you touched on it and you said hurricanes from maine to florida. >> yes. >> and my wife and i had the good fortune of having a house in maine and florida. the insurance in florida is $5,000. the insurance in maine is $500. we're approximately the same distance from the water. i as a native floridian think we're all screwed by the insurance. [laughter]. i can tell you just because i own property in the areas, nobody seems to care. they just keep giving it to us and we keep paying it, and the

mortgage, which i have done, we think that the insurance is terrible. and i would like to know your personal opinion outstanding that destroyed new york city, which, i think, that even a category 4 hit maine, but yet, this just goes totally unrecorded, nobody mentions it and now, all of us floridians pay huge insurance and people in maine and new york pay practically nothing. >> well, it's not clear if that's how the insurance companies work. i mean, they do-- they're highly regulated. at least the primary insurers. well, they are. and that's why they lobby. >> they lobby. >> they lobby and they lobby to the state, so it's largely at the state level. i really don't have a lot of expertise so i could easily say something stupid quickly, if i go down this path.

>> why can't you be more specif specific-- >> well, i do say, you know, you're paying ten times more in florida than you do in maine. i would say you are ten times more likely to get a category 3, a category 3 hurricane in florida than you are in maine. so, just, back of the envelope says your rates are commensurate with at these the relative rates between the two are commensurate with the relative risk. >> [inaudible question] >> but your insurance policy is wind, i'm sure. so, that's what i'm talking about, not talking about surge. >> they're breaking everything down now. that what i'm saying. >> that's-- those are-- those are great points, but i

don't have much expertise on that. yeah. >> [inaudible question] there's a six mile stretch parallel to the coast and five miles in, and responded to six or seven tornados in an eight year period through there and unfortunately, i told my wife, we'll get it one day and our house will get hit. over the years, i'm just wondering the ocean or gulf if there's a zone there with some atmospheric conditions that are coming together during thunderstorms that are getting the thing going. >> that's a great point. there are a lot of local-- i had a figure, i think it's the white figure that shows the tornado the last 30 years in this area. but it's not showing up. but--

>> one night, just about three months ago we were watching a warning coming from sumatra and headed straight-- and they come-- >> you do have kind of a convergence zone here because of the shape of the coastline, so, the area that we call. i don't know what you call it, but it's labeled. >> and that area is kind of a converging zone. so it's possible, yeah, it's possible that you've got -- you've seen in that region more often because you're going to get a thunderstorm there more often. i don't think it's a tornado thing though, i think it's just you'll have more thunderstorms, yeah. but there are a lot of local influences on tornados. in other words, given the fact