Improving Tropical Cyclone Risk Assessment
Chia-Ying Lee is an associate research scientist at the International Research Institute for Climate and Society and a Center for Climate and Life Fellow. She studies tropical cyclones to learn more about their structure and intensity evolution, and how these are influenced by natural and anthropogenic climate change.
Lee received funding from the Center for Climate and Life (CCL) to examine how wind field asymmetries and variability impact tropical cyclone risk and how these can be included in risk models. Tropical cyclones are among the most devastating natural disasters and result in significant economic losses and recovery costs worldwide. Lee’s work will result in improved understanding of tropical cyclone risk, allowing societies to better plan for the impacts of tropical cyclones and minimize losses.
What are the climate questions you’re trying to address?
I am working on assessing long-term risk of tropical cyclones — hurricanes, typhoons, and cyclones — in current and future climates. In other words, I am trying to answer questions such as what is the probability of a given place experiencing, say, storm winds exceeding Category 3 strength in the current and in a changing climate.
In this particular CCL project, and hopefully other future projects — I am applying for funding now with others from Columbia working on this topic — I am trying to address the importance of the surface wind representation to risk assessment. Surface wind, despite its critical roles in causing wind hazard and storm surge, is commonly estimated by using a parametric wind model that generates an azimuthally-averaged field to which a storm-motion induced asymmetry is added on. Surface wind generated under such approach does not cover the full range of the observed wind variability and is, in particular, inaccurate for estimating storms that undergo extra-tropical transition, like Hurricane Sandy in 2012.
The question is whether the asymmetries induced by factors other than storm motion are important to wind risk assessment. If the answer to this question is yes, how can one include these asymmetries? We would like to be able to find statistical relationships between these asymmetries and the ambient environment.
What do you find most exciting about this work?
I find lots of things exciting in this work; I can’t really mark ‘the most.’ The way I approach my CCL work is to first analyze the surface wind structure, by decomposing the whole wind field into mean and asymmetries, in ‘observations’ and numerical simulations. And then I calculate short-term — 10 to 30 years due to the limitation of sample size — hurricane wind hazard by using only mean and mean plus asymmetries, respectively.
It is interesting to see that two datasets give quite different wind hazard estimations. In some places, observational data suggests that we only need to know the mean field with storm-motion induced asymmetries while the model simulations suggest the opposite. There are two possible reasons for that. One is that observations underestimate asymmetries that might be important for long-term tropical cyclone risk assessment. The ‘observations,’ in fact, are analyses with observational information. Observing surface wind is challenging because the traditional remote sensing techniques are not able to see through heavy rain, although there are new techniques that can. The other possible cause is that the model might have biases. Currently, I am trying to explore the possibilities of both causes.
How might this research advance understanding of the challenges posed by climate change?
Understanding the risks we are, and will be, facing is an important step for disaster preparation, prevention, and climate adaptation. A challenging question in estimating tropical cyclone hazard in a changing climate is that we do not have a good answer of what and how a warming climate might affect tropical cyclone activity — landfall rate, landfall intensity, structure, speed, etc. — on a local level. This is because, first, tropical cyclones are rare events, historical data alone is insufficient to estimate long-term risk, especially on the local level, and second, the historical data might not be representable in a warming climate; one cannot extrapolate the results from current climate to the future.
My work tries to solve this issue by generating ensembles of synthetic storms through statistical methods: solving the first problem with environmental input from dynamic global clime models to solve the second problem. In my CCL project, I hope my findings can advance our understanding of the changes in tropical cyclone structure in a changing climate, as well as the associated risk.
What gives you hope?
Humans are resilient. People are aware of the problems associated with climate change and there will be more and more people who acknowledge them. Science is advancing with time. One of the examples is that global climate models nowadays are much more advanced than decades ago. Technology is advancing as well. The renewable energy industries are growing, not as fast as one hopes, but they are growing. Climate adaptation is currently an important research question. Of course, we need more than science and technology. For example, support from politicians can be helpful and it is important to educate the younger generation to respect our planet.
What’s your favorite climate read?
I enjoy reading news from Climate Central. I also like to read news and blogs about extreme weather.