AIR is eyeing models for both U.S. flood risk and for how pandemic risk could impact a population, according to Peter Dailey, vice president and director of atmospheric science for AIR. Dailey, who said the flood model should be ready by 2014, spoke to Best's News Service at the recent Risk and Insurance Management Society's annual conference in Los Angeles.
Q: How is climate change impacting risk management?
A: At AIR, we of course develop a wide range of models for extreme events, and helping insurers and risk managers to prepare in advance for the kinds of extreme events that can occur in a year. When it comes to atmospheric events, which is my area of specialty, climate change is definitely a topic in the front of a lot of people's minds. For risk managers and insurers, to some extent, climate change is a question in their minds. "Is this something we need to pay close attention to for our portfolio of risks?"
What we do is we study climate from all angles. A full spectrum of impacts it could have, all the way down to individual events, the effects of climate conditions on, for example, hurricanes as they're developing in the Atlantic, and potentially impacting a portfolio of properties in real time. We've just recently announced a service that helps insurers and risk managers to account for that risk as storms develop.
All the way through the other end of the spectrum, which would be looking out 20, 50 years from today. The uncertainties are quite large, but it may be that climate will change to such an extent in the long term that even insurers and reinsurers need to be carefully paying attention to those types of impacts today.
We also study the long-term perspective, we look at how climate conditions may develop and evolve over the long term, and affect the risks of tomorrow.
Q: What emerging risks are you seeing today?
A: We've been seeing a lot of different interests here at RIMS. Those include some areas of risk that we're developing new models for, for example pandemic flu. This is an area of emerging risk that's becoming more and more pertinent. As a modeler of extreme events, we're looking into how very widespread pandemics could affect very large populations, whether it be in North America, or anywhere in the world. We're also hearing a lot about the need for objective assessment of U.S. flood risk. There's not any one single model that can account for U.S.-wide flood risk today, so this is something that we've been developing, actually, for some time now. It's a very important area for U.S. insurers, obviously, re-insurers as well.
This is a model we've...It's been in the works for quite some time now. We expect to release a U.S.-wide flood model for the contiguous US in 2014. That's another area of risk that we've been hearing a lot about here at RIMS, and we expect to move forward with more research.
Q: Looking back at Hurricane Sandy, what lessons were learned there?
A: Sandy was one of many very important extreme events in the very-recent past. Ranks up there with other major recent events like Hurricane Katrina, for example, the North Ridge Earthquake in California. Every time there is an extreme event, whether it be in the US, or anywhere in the world, we pay very close attention, not only as the event's unfolding, but after the fact. We need to take a close look at the kind of damage that's observed on the ground.
We take a close look at how well our model was able to simulate the circumstances of the event, whether it be the hazard, the wind speeds in the case of Sandy, the flooding along the coast, or the actual damage that can occur to property. How widespread that damage is. Some of the lessons that we're hearing come from domains beyond just insurance. For example, emergency managers are very interested in how Sandy unfolded, but also how can they better prepare well in advance for an event like Sandy, before it happens.
This is also, obviously, of interest to the insurance industry. One of the reasons the cat-modeling industry itself is around today is because of the aftermath of Hurricane Andrew, where many lessons were learned. What we find here, especially in the case of emergency management, is that by playing out scenarios in advance of the season, could be a historical event like Sandy, it could be an even more extreme event than Sandy, but a plausible event. Maybe more extreme than has been observed historically.
What you find is that decision makers are able to play out an event, make decisions, but without the stress of the actual event unfolding. In advance, learn from their decisions, what kind of information would allow them to make better decisions. That kind of playing out historical events, reconstructing historical events, and putting decision makers to the task of decision making in advance, allows them to better prepare. That's what we're in the business of doing developing models that allow any industry of interest in risk, to better prepare for events before they happen.