Rain Men


Among a handful of powerful initiatives designed to confront flood hazards head on, FM Global has assembled a highly accomplished flood research team that will focus its efforts on the large-scale mapping of critical flood zones.

How did you first get into the field of study related to flood, mapping and modeling?
Dr. Jeff McCollum, senior research scientist:
I was in engineering in college and one of my civil engineering professors was a big flood researcher. Through him I got into the field as an undergraduate, which led to my professional work in civil engineering and flood research. I worked for the National Oceanic and Atmospheric Administration (NOAA ) for seven years where I studied remote sensing of rainfall. After seven years working for the government, I needed to try something different. This job opportunity surfaced on a National Aeronautics and Space Administration (NASA ) email list. I was looking for a more applied occupation, which is more practical and efficient. Plus, a lot more gets accomplished in the private sector!

Right now I’m in my seventh year at FM Global and I mostly focus on flood, but also rainfall that causes flood. My first mission is to understand how big a flood could be, because we don’t want to be surprised by a huge flood event. Before a flood, and after a flood, we compare the data we had for expected floods to make sure we were on target.

Dr. Yizhong Qu, senior research scientist: I majored in hydraulic engineering and also studied geomorphology, which is the study of terrain. I moved to Pennsylvania State University to study hydrology, which is a field somewhere between geomorphology and hydraulic engineering.

After graduating in 2005, I began working at AIR Worldwide, doing some catastrophe modeling. While I was there, I developed a U.K. flood model, which included large-scale storm simulation.

I joined FM Global in November 2010, and am doing similar work, supporting field engineers in large-scale engineering and site-specific work. Together, we are developing solutions that are suitable to FM Global’s clients.

Dr. Xinya Ying, senior research scientist: I studied hydraulic engineering as an undergraduate student and in my master’s program in China, where I became a research engineer. I earned my Ph.D. in computational hydroscience, which is basically numeric modeling, in Japan. In 2001, I began working on flood modeling and dam breaks, doing one- and twodimensional modeling needed by the U.S. Army Corps of Engineers. I arrived at FM Global just a few months ago to join this new flood team.

Dr. Ed Beighley, senior research specialist: I received my undergraduate and master’s degrees in civil engineering, with a focus on hydraulic engineering. I created terrain models of the earth. I received my Ph.D. in civil engineering, where I combined remote sensing, geographic information system (GIS ) and hydrologic modeling. Then I did a post-doctorate at University of California, Santa Barbara, where I modeled terrestrial run-off to the ocean. I taught GIS and hydrology at San Diego State University, where I became interim chair of the civil construction and environmental engineering department.

I put a lot of time and effort into developing a hydrologic model, which won a NASA New Investigator Award. The model allows us to go anywhere in the world, using remote sensing, to set up a river network and model rainfall run-off processes and streamflow with limited in-situ data. This provides a basis for what bigger flood events might look like.

Realistically, I felt I was going to be a professor forever. I had been working with FM Global for three years via contracts to do some of their modeling work. I was looking at characteristics of what causes a place to flood. Every time I’d come to visit FM Global, I was so impressed with the level of expertise, the commitment to research, and the weekly Research seminars. When the opportunity arose to work here, it was too good to pass up.

What would you like to accomplish individually and as a team?
We all have a different skill set. The biggest thing I’d like to see accomplished is to integrate larger scale hydrologic and climate models, which would involve working closely with Jeff to get the climate input right into what I do.

Ying: I’d like to couple my one-dimensional and two-dimensional hydraulic models with our hydrologic models, improving their efficiency and speed to handle large-scale river networks and watersheds for flood risk evaluation and for flood mapping that can consider dam, levee and other flood defense failures.

Qu: I would first like to provide solutions to evaluate the risk of flooding. In the past I have developed an integrated hydrological model. We want to make sure it works. Conceptually, it’s correct and technically it’s advanced. But right now we haven’t focused on solving the real problem, and on providing a reliable solution for the future. Together we can develop a package of solutions, good for both large-scale mapping and for very detailed, site-specific applications.

McCollum: My work will be more on the precipitation side. Big floods, the kind we’re talking about, are usually caused by precipitation. What could happen? Why would huge rain events occur? I would use statistics and climatic data that could be put into the modeling that my colleagues are all using. Then we could find out directly about the vulnerability of our clients’ locations.

What are your thoughts on working together with this team?
In our field, most people would focus on one aspect of the equation and rely on third-party data for the rest. But if we put everything together—meteorology, hydrology and hydraulics—it would be unique. Only a few other firms do it and most are simply in the business of catastrophe modeling. We are looking for where the specific risks are, throughout the country and at our clients’ individual sites. We care individually about each risk. Working within a team is not new for me, but with this ambitious goal, it’s quite unique.

Beighley: We have a unique opportunity here. As a professor, I managed a team and managed many projects at once. Now if you look at the people in this group, individually we could all serve as the top of a team, or do something completely independently. But to get to work together, with such a broad category of experts so closely is really rare. To have a team where the individuals are so qualified and don’t have to manage other teams, that’s pretty unique. Normally, people in our capacity would spend a lot of time searching for funding. At the university, I spent much of my time on finding research funding. Did I get to solve the problem defined by the project? Hardly. That’s the big difference. We get to solve the problem here.

So what is the first step now that you’re all here?
Flood mapping and modeling is a corporate objective. From a “research community” modeling perspective, we had advances in hurricane or typhoon modeling in the ’80s, earthquake in the ’90s, and now it’s flood modeling in the first decade of the 21st century. But much has changed. Computational power is the biggest one. Flood modeling requires a lot of data, and collecting that data was impossible 20 years ago. I can see that flood modeling is possible right now because of these advancements. At the same time, there are quite a few significant events, which will make it challenging as well.

Beighley: We’re all here. We have this idea of what we want. Luckily, we are all able to do what we are passionate about. As the starting place, what I see happening is to figure out how we all fit together. We’ve got experts in precipitation, hydrology and hydraulics. If we can pull together, pull out the contributions from everybody, and put it on a single slide that we can discuss, it will all come together.

Ying: First you must predict how much rain comes to the ground. Then you must predict how much of that rain will discharge into the river through hydrology. Then you need a hydraulic model to calculate elevations and flood to create the flood map.

Do you feel that there is urgency on the flood team to produce results based on the kinds of natural events we’re seeing so often?
Well, as the population expands, people always want to be by the water. Even when there’s a flood and places get ruined, the cities want to have their downtown areas on the water, and therefore, more and more property gets exposed to flood. Clients might consult with FM Global before building, but most will still locate there. It just puts them at greater risk for loss.

Qu: The amount of exposures inside the flood plain is growing rapidly. So, nowadays, the same event today versus 10 years ago will cost a lot more.

Ying: The fact is, cities that usually have large impervious areas are growing and rain cannot infiltrate into the ground, producing large run-off, which the drainage network cannot accommodate. This will potentially expose the cities to significant and very costly floods and flood damage. But also, people tend to forget, especially if flood risk is not communicated properly. Last year’s flood in Nashville (Tenn., USA ) is a good example. Most people didn’t anticipate that, but it did happen and it was very damaging.

Beighley: From an engineering perspective in terms of flood, there are two lines, the 100- year flood plain and the 500-year flood plain. They are literally two lines on a map: on one side you can get this insurance coverage, and on the other side, you can only get this coverage. Right now we’re trying to get a handle on a consistent way of looking at these things, to really understand why these lines are drawn and what the uncertainties are.

Moving forward, will any prospective change in climate have an effect on what you do?
Yes, I think it will. It’s hard to say. The climate is changing. But the question is also what’s causing it? Together, we pay attention to all of the climatic data sets.

Beighley: It is a challenge. In certain areas, for example, it may be more intense here but not there. There’s more rainfall in one place now than there used to be. Individual events might change for whatever reason. But when you push these changes out and consider the estimation of 100- and 500-year flood elevations, that’s when the uncertainties become greater, because the events are so much more extreme. In any event, we have to consider that there may be more precipitation, and, historically, this may alter the magnitude of flood events. There may also be more winter precipitation, potentially more snow in places that didn’t have that much snow. To the extent that the 100-year or 500-year flood elevations may be impacted by this snow melting will be hard to quantify. It all remains to be seen.