Gulf oil spill + Most active hurricane season ever = ?!?
Of big concern in the Gulf of Mexico right now is, obviously, the BP oil spill. What residents along the shores are usually concerned with at this time of year, however, is hurricanes. Katrina is still fresh in the minds of many Louisiana residents, and indeed the 2005 Atlantic hurricane season stands as the most active season recorded for the region.
Most active season, that is, before 2010.
NOAA is expecting this season to steal the throne from ’05, which is enough to cause a big pile of worry. Worse, though, is the effect any one of the 14-23 anticipated storms (3-7 major hurricanes) could have on the still-gushing oil spill. I think Randall Munroe over at xkcd has a pretty good take on what this will look like.
To try and get a head start on planning and preparing for the soon-to-ensue natural calamity, a set of congressional briefings was organized by the Natural Hazards Caucus - one on the House side, and one on the Senate side – to try to keep the policy-makers in the loop.
I went to the House side briefing, and the story-proper will be up soon-ish (*Edit: the post is up now) – but in the mean time, for anyone who finds themselves exceptionally curious/bored, I have written up a transcript (and there are links to the slideshow presentations!!!!) of the four speakers. There are also audio recordings of the presentations, so you can follow along with the slideshows and pretend you were really there!
Heidi Cullen – CEO and Director of Communications for Climate Central was the moderator of the session. Heidi gave a brief overview of the topics to be covered, and made some interesting points about the Loop Current.
Greg Holland – Director, NCAR Earth Systems Laboratory, National Center for Atmospheric Research. Greg spoke about why the upcoming hurricane season is expect to be so bad, and how climate change plays into hurricane formation.
Rick Leuttich – Professor & Director, Institute of Marine Sciences, University of North Carolina at Chapel Hill. Rick was on hand to demonstrate his model [like the one shown above], and explain how hurricanes will interact with the oil spill.
Rowan Douglas – CEO, Global Dynamics, Willis Re and Chairman, Willis Re Research Network. Rowan talked about the history and importance of public science in the insurance industry.
The transcripts are after the jump.
Welcome everyone. It seemed like the day that hurricane Alex will be making landfall in Northern Mexico as a cat-1, potentially category 2 hurricane is probably a very good day to gather a panel of experts to talk about the risks of hurricanes and how they would impact the oil spill in the Gulf.
I want to start off by thanking our sponsors… [here]
So we’re going to kick off a discussion with three experts… Greg is going to answer some questions about the “Why?” Why is this hurricane season expected to be the most active, potentially, on record.
… Rick is going to focus on the “How?” How an oil spill might interact with hurricanes if they do form and cross the Gulf of Mexico.
And then finally we have a reinsurance expert with us here [Rowan]. [These experts] are going to help us put these risks into perspective and assess which parts of the Gulf, and the states along the Gulf coast are most at risk, as well as which states along the East coast will be at risk.
To start out I just want to frame this discussion by giving us a moment to look back at the 2005 hurricane season and to say that – right now we are dealing with a worst-case scenario, the Deepwater Horizon oil spill is a worst-case scenario. And I would say that the fear in the back of many of our minds right now is – we could potentially see the repeat of the 2005 hurricane season. We want to help folks walk through what that double-whammy of worst-case scenarios could look like, what it could mean, and how we can prepare and manage those risks as we head into the peak of hurricane season and through the fall.
Needless to say the 2005 hurricane season was the most active on record: 28 storms, a record $128 bn in damage. So, I want to more-or-less set the table as to what some of our experts are going to cover. Again as I said, right now NOAA’s forecast is for the most active, potentially, hurricane season on record. Right now they are calling for 14 to 23 named storms, 8 to 14 hurricanes (top winds of 74 mph or higher), and 3 to 7 major hurricanes. With Alex in the Gulf right now, we can begin to think about how that could impact operations to remediate the oil spill.
Greg is also going to focus on the latest research on how climate change could impact the intensity and frequency of hurricanes.
One reason why we’re worried about a worst-case scenario, and why we wanted to have this discussion today is because when you look at the warmth in the Gulf right now, the warmth in the hurricane formation region is a [large increase] from average temperatures. When you think back to May 2005 we know that one of the reasons the hurricane season was active was because of warm ocean temperatures. In May 2010, the sea surface temperatures are actually higher than they were in 2005, and Greg will explain what that means and why that’s just one of the reasons why the forecast is for a very active hurricane season.
Rick is going to talk a lot about what a hurricane interacting with the oil spill could look like and he’s going to focus on where the risks would lie given the trajectories of different hurricanes. I also just want to add that there is another aspect to the hurricane season, and that is the Loop Current, because one of the secondary questions that’s come up is – what would happen when the oil gets into the loop current which is literally a current that connects to the Gulf Stream, and would bring oil potentially up and along the Eastern seaboard.
Right now I think the most important update to give is that until recently, the loop current was essentially broken. We had an eddy, called Eddy Franklin, or the Loop Eddy, which was literally blocking the oil’s entrance into the loop current. As of last week, that eddy had begun to show signs of reattaching which essentially meant that oil could now begin to interact with the loop current. What that tells us is that there is also a certain amount of risk associated with oil making its way into the Gulf Stream.
If the oil does make its way into the loop current, the timeframe we’re thinking about is about one-to-two months to move northward into the Gulf Stream. The Gulf Stream essentially breaks off at Cape Hatteras so the risk would essentially bear out up until Cape Hatteras.
So a basic summary before I pass it off to Greg:
The 2010 hurricane season is expected to be highly active.
The highest oil risk right now – western Louisiana to Texas as well as the marshes east of the Mississippi Delta. And, lower but significant of oil getting from the loop current to the Florida current, and eventually into the Gulf Stream on a slightly slower trajectory.
I’m going to talk about the potentially disastrous 2010 hurricane season – and those words are chosen quite carefully. There is enough information now that we have to actually plan on this being something like a 2005 season – it may not happen, indeed I sure as heck hope it doesn’t happen, but it’s at the stage where you just simply can’t neglect it. That comes from a range of reasons including environmental and ocean temperatures.
[The 2010 sea surface temperatures for May, when compared to the 2005 sea surface temperatures] – If you just look at the patterns you’ll see that they’re actually quite similar. We’ve got a very warm ocean out in the eastern Atlantic, not quite so warm ocean in the western Atlantic and into the eastern Pacific. The big difference is that the 2005, which was the previous record for the eastern Atlantic, is about a degree to a degree-and-a-half cooler than what we’re seeing at present. So we’ve got the same pattern, but it’s actually at unprecedented levels.
That means that the environmental changes that occur in relationship [with the warming]; the weather patterns, the way in which the air circulates around the Atlantic Ocean is also changing. And because the ocean is one of the major drivers of the atmospheric circulation patterns we are actually starting to see quite similar circulation patterns appearing as happened in the 2005 season.
This is something that, if it’s happening now, the chances of it going away are unlikely – it’s not impossible, but they are quite unlikely because we are actually now locked into a cycle which will continue almost certainly all the way through the hurricane season.
So we have the season hurricane forecasts starting to come out. NOAA has estimated 14 to 23 named storms as Heidi just said. There has not been a seasonal forecast of 23 storms put out for this country before. Now that’s over the past 20 years when seasonal forecasts have been made, so it gives you an idea of how the community is reacting to this.
All of the other seasonal forecasting schemes and predictions are in the same range, some are a little bit lower, some are a bit higher.
So we have easterly waves coming off Africa and they are one of the major formation processes for hurricanes in the north Atlantic. They are coming across that warm water and there is a lot more energy. It’s sort of like putting a turbocharger in your car – the more efficiently you can use the fuel, or the more fuel there is, the more powerful the system can become. That water has also changed the circulation pattern. So there is lower pressure in the tropical region, reduced vertical wind-shear (the change in winds with height, so that the storms don’t get tilted out and destroyed by the vertical wind-shear), and all of the right patterns to both form more cyclones, but perhaps more scarily, that region out in the eastern Atlantic is where we get the infamous Cape Verde hurricanes, and they’re the ones that are almost entirely responsible for the really damaging storms.
So what’s happening? Well it’s a combination of global warming and natural variability. [The global mean surface temperature, and the mean temperature for the hurricane forming portion of the Atlantic ocean have increased.] May 2010 was the warmest global temperatures at the surface that we’ve ever observed, and May 2010 in the eastern Atlantic is the warmest it’s ever been. So there is a trend. But you can also see that there is a lot of natural variability [on both long and short timescales].
We’re actually in a high period on a decadal scale of activity – so we’ve got a trend with a high period sitting on top of it. And finally, we’re actually coming out of an El Nino – the El Nino was responsible for the substantially reduced hurricane season we had last year. We’re coming out of that, which means we’re going into a more active year, even compared to last year. The forecasts are now predicting a La Nina, which takes us even further into that scale. So this is why everyone is concerned.
[The warming Atlantic] provides the changes in the circulation patterns and it provides the energy – and you simply can’t get away from that – there’s randomness to the formation and development of hurricanes, but if you set up the conditions it doesn’t matter, the randomness will eventually take over.
You really do need to take this in the context of a longer scale, and the outlook for Atlantic currents and climate change is that you need to start expecting more of this. If we look at just annual frequencies by themselves, purely for the Atlantic ocean, there is a lot of uncertainty. Some studies are predicting more, indeed some are predicting 20-30% more, some are predicting less. So we don’t really know how the future climate change will affect hurricanes, except for the fact that they won’t get substantially less, and they may get more.
But there are very consistent, almost uniform predictions coming out which are showing the mean intensities increasing by about 5-to-10%. And you might say, “Well okay so what? So [the wind speed] goes up a few miles-per-hour, that doesn’t really matter.” The unfortunate thing there is that a 5-10% change in the mean leads to a 50-100% change in the major hurricanes. There are no good stories here, it’s just simply the way the system works out, so we’re looking at potentially a doubling of major hurricanes in the next 20 to 30 years.
One thing that often gets overlooked is that rainfall is consistently predicted to increase by about 20%, and you might recall that the worst hurricane disaster we’ve had in the last 50 years in the North Atlantic was Mitch, which wasn’t even a hurricane, it went ashore as a tropical storm in Central America. All of the damage, and the several thousand people who were killed was entirely due to rainfall. So we need to do adequate planning, and I urge you to start thinking these things through.
Finally, are we prepared for this season?
Well in terms of hurricane tracking, we’re doing pretty well. Research over the past 20 years has actually reduced the error – so if they say the track is going somewhere, there is a very good chance that’s where it’s going. Indeed the errors now for 3-day forecasts are actually about the same as they were for 1-day forecasts 20-odd years ago.
In terms of intensity, there’s been no improvement in the forecast at all. There is a major program going at Princeton, and there needs to be a lot more work done on this, but there’s been no improvement. That means that because we say it is a certain intensity, you have to take into account that it could be less, and it could be more.
[For tools that are used in decision making], we’re moving from saying, “There is a cat-3 hurricane coming into some region,” to, “This is what will happen to the region.” There is a rapidly growing community which is focusing on this.
The question I want to ask you, just to give you an idea of how unintuitive a lot of these things are – Which parameter would lead to the most damage to offshore energy facilities; the intensity, the size, or the forward speed of the hurricane?
I think a lot of people would say intensity, a few people would say size. It actually turns out that the forward speed is the most important thing, a slow moving storm does enormous damage compared to a fast moving one. Size is next important.
If I actually only know the forward speed and the size well, the intensity hardly matters. It doesn’t matter if it’s a cat-2 or a cat-5, if you’ve got those first two parameters you’ve got almost all of the damage that occurs in the off-shore energy industry.
So to summarize; there are no good indications for the 2010 hurricane season – Please be prepared. I sure as heck hope it doesn’t happen, and I’m more than happy to be proved wrong, but at this stage you have to start planning.
The climate outlook indicates that this is more of what we’re going to be seeing. There is in particular a sustained potential increase in major hurricanes. And finally, the decision tools that we’re starting to develop will actually make it a lot better for you, and for your constituents, and for industry, to actually be able to understand what the real impact of a hurricane is instead of just the meteorological parameters.
If you’re here today, I know that you probably have a reasonably good idea of where the oil is in the Gulf of Mexico that we know about. It’s all over the internet, it’s what you see on TV. Most of it is east of the Mississippi river delta, and it stretches over to the Florida panhandle, and there’s a big blob of it that’s impacted the shorelines in those areas. The main message of what I’m going to tell you about today is just that hurricanes can be real game-changers, and expose areas that really have thus far largely escaped impact from the oil, and I’ll you how and why that is.
There are three main consequences that I’ll point out about how oil and hurricanes will interact. [But first] for the most part I think we believe that oil isn’t going to change hurricanes very much, it’s that hurricanes are going to impact the oil.
One [consequence] is maybe a good thing, and that is the high intensity winds and the energy associated with the storms will do an awful lot of mixing in the water column. It will cause some dilution and dispersion, and may help with the degradation process. So from that perspective, that’s a good thing.
The second consequence has to do with the way in which hurricanes work. The winds cycle counter-clockwise around the storm, and consequently the wind on the right-hand-side [of the hurricane] will push water in the direction that the storm itself is headed. And on the left-hand-side [of the hurricane] it will pull water and anything that is associated with the water away from the direction the hurricane is headed. So you can see that depending on the direction of the storm, and where it makes landfall, you can have [different effects]. [On] the right-hand-side [the water will] push onto the shore – and if there is oil in those particular areas you can expect that oil to inundate the land. The left-hand-side [of the hurricane] will pull offshore. It is very dependent on the direction and the location of the storm.
To remind you of what that might mean in the context of four of the major hurricanes that have hit the Gulf Coast in the last 5 years – Katrina, Rita, Gustav and Ike - [he has graphics of what it would have been like if the oil spill had been in place when these hurricanes had gone through. They can be seen here.]
Any of those areas where oil is proximal, and right now we know that is mostly the Mississippi and Alabama coasts, depending on where the storm goes there’s going to be oil in contact with land in those areas.
The third consequence, and in some respects the most interesting, is what happens not onshore or offshore, but along-shore. In each case, you’re still going to have counter-clockwise rotating winds, and during some phase of the storm you’re going to have winds that are basically shore-parallel – That are counter-clockwise but are going from east-to-west in the Northern Gulf, from north-to-south along the Texas coast, and from south-to-north along the Florida coast. It really doesn’t matter where the storm goes in the Gulf, you’re still going to have those general persistent winds.
The closer the storm is to the US mainland the stronger it will be. But also the larger the storm is, the more intense the storm is, the stronger it will be. But even hurricane Alex right now, which is way far in the southern part of the Gulf, is still generating along-shore east-to-west winds along southern Louisiana.
What this does is it drives water, and anything associated with it, from east-to-west in a counter-clockwise motion around the Gulf. It can do it very strongly, and this can be the game changer I mentioned at the beginning.
So largely the oil has been around the Delta region, and [it has been] the Mississippi, Alabama, and western Florida coasts that have been impacted. So we have used a very sophisticated hurricane model, and then seeded that model with particles that represent pseudo-oil in roughly the area that the most recent imagery shows there to be surface oil.
As the [modeled] storm builds up, you see a very strong east-to-west counter-clockwise transport of the particles around the Gulf and consequently a major displacement – the particles make it all the way around to the Texas coast down to around Galveston. So if a major storm comes along there will be a very substantial redistribution horizontally of where the oil is in the Gulf.
As you make that counter-clockwise rotation around the Gulf you push things to western Louisiana, you push things to Texas – areas that haven’t seen them. But you also push them up against the Mississippi Delta, and so those marshes just east of the Mississippi Delta now, which have largely been un-impacted, they will see a major push of materials that are [as of right now just] sitting right offshore.
To summarize; this counter-clockwise circulation – the strength of it will depend on where the hurricane goes, but the fact that there is a strong hurricane in the Gulf will generate it almost regardless of where it goes. It will have an impact, and I think this will be one of the most significant [impacts] on the oil.
Again, hurricane Alex is out there now. It is having this same sort of impact, and our most-recent forecast models are showing a net east-to-west movement, particularly of oil that has up until this point been near the mouth of the [Vera Terra???] basin. It’s pushing it to the west towards the [Chaffeleya???] as well as, to some extent, into the marshes.
So to end – three consequences: The mixing may be a help, the counter-clockwise winds on the right-hand-side push the oil onshore, on the left-hand-side pull it off, and that is very highly dependent on the track. And then lastly, but not least, these counter-clockwise winds also push water in a counter-clockwise motion around the Gulf, and I think it will likely expose areas which have largely escaped the direct consequences and impacts of the oil.
I work for a reinsurance broking company which is involved quite heavily in public science. I suppose I sit in the shatter-zone between catastrophe, particularly hurricanes, and capital. I’m not going to talk perhaps quite so much about the science as my learned colleagues – I’m going to talk a little bit about having to confront decision making in a regime where the past is increasingly unlike the future, and that has profound implications for social, economic, and I dare say political decision making, but I’ll leave that one up to you.
The fundamental challenge we have is that this hurricane season that we have ahead of us, confused even more by the oil, represents a remarkable case-study of how we manage extremes in a modeled world. Whether it’s financial extremes or environmental extremes, we have to try and make decisions based on models, and science is underpinning that.
So if I was to frame the coming hurricane season within a social and economic context, this is the grand challenge for global society – How do we understand and evaluate the costs of extreme events and share those costs at local and global scales through public and private mechanisms? How do we make the decisions to allocate resources to mitigate or adapt or create these solutions? Nowhere more-so is this the case than in the U.S., and nowhere more-so than in the case of hurricanes.
That stimulates, in the case of the hurricane season every year, three fundamental questions. The first is; “What is the maximum probable loss, both economic and social, that we can confront in a typical hurricane season with a one-in-two-hundred year return period?” That’s a fairly strange phrase I’ve just used, and I’ll explain it. Around the world if you buy an insurance contract, and I think the convention is much the same here in the United States [as it is in Britain], regulators and policy holders believe that that insurance contract, that promise, should work 199 years out of 200 – so insurance contracts should be able to respond to the worst probable loss at that return period. That’s the first thing people worry about.
The second thing people worry about is; “Okay, what is next season going to be like?” And then some people, particularly regulators and public officials say, “And what’s the average?” So those are the basic questions.
Now, if you look at land-falling hurricane activity over the last 100 years it goes in waves, these natural variations, but there hasn’t been a tremendous change from the norm yet. We worry about the future. But the level of losses has been absolutely vast in the last 20 or so years. This profound economic impact has had huge consequences, not least of all on the industry that I’m part of.
[The losses to the insurance industry since 1970]… is a history of the role of science in finance, or this side of finance. [Describing a graph] Wind storm losses tracks pretty normally up to about 1985, then you’ll see a first little peak, it’s not actually labeled, and that is Hurricane Hugo. If you’d been under-writing these risks through state schemes or private ones, this was a massive event. And suddenly people thought, “Is this a one-off? Or is this a big change?” And some people thought we should analyze things in more depth, and they invented the spreadsheet and all sorts of things happened which we forget about.
People started saying, “Well how many storms could happen? How strong are buildings? How vulnerable are they? How much economic loss? How much insured loss?”
Then in 1992, Hurricane Andrew came along, whizzed through Peurto Rico, hit Charleston, tears before tea time. Many insurance companies went bust, lots of social dislocation, lots of people thinking, “What do we do next?” and there was a huge investment in public science and modeling.
In 2005, hurricane Katrina, followed by its cousins Rita and Wilma, an unprecedented level of loss – the insurance industry creaked, very few companies went bust, and it was a tribute to the power of modeling.
Interestingly, much of that insured loss, the £50 bn ($75 bn USD) wasn’t paid for by Americans, because most insurance companies are reinsured for these sorts of events. That is a concept that is shared globally. So most of the payments for those people in Louisiana and elsewhere – of the insured losses – came from outside of the U.S. Just as well, for a bush fire in Australia, or a wind storm in Europe, or floods in the United Kingdom, you, if you buy insurance, will probably be contributing to that. Insurance is the ultimate community product, perhaps second only to taxation.
The next stage was the real embracing of public science to help us understand extremes we haven’t confronted before, and that’s spurred a huge development.
Going to Europe quickly, financial regulators of insurance now use models to capitalize insurance companies for the one-in-200 year return period. Natural catastrophe drives this. If you change the models, ie: new versions or new inputs, the capital requirements for insurance companies can change by hundreds of millions of euros or dollars. So this means that modeling in science is becoming a combat sport.
Here is some modeling work on how much a category-3 hurricane in New Jersey and New York would cost. There are nine variations of how these models can be constructed, but the selection of these models can change the estimated economic damage of a cat-3 in just these two states by about $40 bn.
This is the challenge we’ve all got as politicians or as business people or as academics, we’ve got 60 years of observed data in the Atlantic at best. The rest of the world has much less than the North Atlantic. How on Earth do you work out what is a one-in-200 year maximum probable loss with only 60 years of data? You can take 60 years and use some clever statistics to broaden that, but fundamentally we’ve got a tiny dataset, and the climate is changing. Was 2005 a one-in-200 year event? We’ll we could have another one now. We’ve had the hottest world May temperatures ever, and now it’s the hottest American month of June.
What we are now able to do through the power of public science is simulate the planet in ways that have not been possible before. We now have at our disposal [the possibility of taking] 300 years worth of current climate conditions, model that and look at how many tropical cyclones come through, and use that as our base. So we took these 300 years, we found six years with five significant category-5 hurricanes with landfall, [and the worst three year period was very recent.] This is just to say that we’ve got to start investing further in the science to help us understand these questions.
You go from the global to the local, this is work that is being done at NCAR to really refine our understanding of the behavior of tropical cyclones as they hit landfall in the Gulf and eastern seaboard region. [Then we can] take those global models and understand the frequency [of these event.] This is blue-sky public science, but what we’re seeing now is a new supply chain. The data and information from public science is immediately going into the models which are used by financial institutions and others, which is then influencing public policy decisions. We’re seeing a new supply chain which is absolutely critical, and the masters of the modeled world will be those countries who are at the leading edge of it.
So similar work that we’re involved with is trying to understand how much flooding will occur from tropical cyclones. As Greg indicated, it’s actually rainfall that can cause the biggest challenges rather than the winds and the storm surges.
Taking the work that Greg was doing at NCAR of understanding what actually causes damage, moving from hurricane hazard through to loss both in off-shore platforms and on the mainland, and this is creating a new index which is much more accurate. This isn’t about understanding hurricanes better, it’s about liberating capital, new markets, to be able to come and provide solutions to those who will be put at risk.
[This next slide] is really important, because everyone is worried about what will happen this year or next year – In Geneva in September there was a World Climate Conference-3 meeting. The first one I think began the IPCC, the second one set up the Kyoto Protocol, the third one is an intergovernmental agreement to drive global climate forecasting services, going beyond our two-week horizon through to seasons and perhaps 18 months to two years ahead. Among the two leading players in this are the Americans and the British – this is going to be critical to operationalizing proper forecasting using these new computer models to have the same sort of confidence about weather forecasts as we do about climate forecasting.
Many questions remain as we try to work out the risks we face from hurricanes. This is going to require a partnership between the public and private sectors to take forward these sorts of questions. But the big movement that is occurring is a re-definition of what sustainability means. Sustainability is a fairly mushy concept, and most people feel as if it is hard to get your hands around. Actually, sustainability – whether you’re worried about financial markets or environmental extremes or other areas is about understanding and managing extremes within tolerable parameters. Trying to [prevent] extremes [from occurring], and if they do, ensuring individuals and societies are resilient [enough] to cope. This is an object lesson in making states or regions resilient to one-in-200 year return period events.
It’s a fascinating challenge we have, and one that I think will unite us all.
[Most of the Q&A was pretty dull, but this one question was interesting]
Q: Does the model take into account sub-surface oil?
Leuttich: … [People who try to model the deep-ocean] are given an incredible challenge because it’s so hard to see anything down there so they don’t necessarily know where the oil is to start with, they don’t know if their models are actually correct because there is so little data coming out of there. But nevertheless, they are tasked with and probably are as good as anyone at anticipating and giving science’s best guidance on where the oil is going in the Gulf proper, the larger areas of the Gulf.
The work that we specifically do doesn’t address that, but rather it addresses the near-shore areas, and those are the areas – Hurricanes obviously do come across the Gulf, they do mix the water column in the Gulf, but they tend to only mix down a couple hundred meters or so. Much of the oil that is deep in the Gulf is probably below that mixing level.
Now once you get up on the continental shelf that’s a different story, and [a hurricane] will mix the entire water column. If oil gets down and settles out in the sediments it may comes back years later if a storm comes through and stirs all of that up. But the work that we specifically do, where the storm surge occurs, where the real impacts of the hurricane occur as least in a societal sense, are in the near shore. So in these events you can anticipate that what is on the surface will probably be mixed through the water column, and it will move along with the surge.