WT: Professor, you are doing research around the FEMA floodplain maps that apparently are quite outdated. Can you tell us what your role is in this, and, how bad it get with these FEMA maps?

Sankar Arumugam: The FEMA maps do depend on USGS information for flood frequency curves.They do update the maps, but that update needs to be done in coordination with state agencies. This process is slow, (as) it naturally requires more coordination.

There are two parts to the challenge: one is revising the FEMA maps with current information. The second challenge is projecting changes in flood frequency based on future climate projections.

This is where the challenge comes in. The first gets updated slowly, but on the other hand, the part related to revising the flood frequency curves, and predicting the flood frequency curves based on climate information is a bit more challenging, as the methodologies are evolving.

WT: That’s a fantastic amount of work. How long, would this work go on in perpetuity because this is such a large project?

Arumugam: This project we are working on is in coordination with the scientists from the US Geological Survey. In the US, if you want to design, for example, a spillway for a one-hundred-year period, we use the map and the flood frequency curves associated with it. The paper published last week proposes a different approach; as opposed to using hydrologic models, using a statistical approach for estimating flood frequency curves thirty years ahead. The most important thing is that we provide calibration and provide validation of the models for the past twenty and fifty years.

One of our findings is that the challenges are more in the arid southwest, where it is much hotter, because the variance of the floods is a lot higher, compared to the more humid areas like New England. Where it is more humid, this work is less challenging.

We got two more projects in terms of rewriting these frequency codes, working with various agencies, including the USGS, the Dept of Highways and the Army Corps (of Engineers). We need to work with them not on the methodology but on the challenges of applying flood data/projections. The project collaboration with these agencies, particularly the National Science Foundation (NSF) is going to be for another three years.

WT: According to the Southeast Climate Adaptation Science Center you are leading a study on assessing the impacts of hydro-ecology over the sun belts in the near term. How do you see climate change from ten years ago to now affecting all these types of studies? I hear about using artificial intelligence combined with normal modeling. How does it work, given what you do, climate change predictions done on a large scale, and how does AI help you? Give our viewers a practical example.

Arumugam: One well-known aspect in the context of climate change is that the extremes are more frequent, and the variance of normal precipitation or temperature that we are experiencing is also changing a lot more, varying a lot more.

The approaches that we use here are two-pronged: one uses a physical modeling approach, and the other a statistical modeling approach. The way we do it is to use the best of both, a hybrid modeling approach.

For example, if you want to estimate the changes in precipitation, I think you need to use a hydrological model. On the other hand, if you are interested in an upstream basin with limited human influence, I think you need to go with a statistical modelling approach.

We choose the approach depending on the context. It is not like one size fits all.

We need to look at multiple models in terms of the projection of what kind of precipitation and temperature are available and use that information to quantify the changes to the flood curves, and the implications for the ecology.

WT: You were involved in an interesting research paper, the Urban Multi-plex. Can you tell me what makes this concept so important for people that would do flood management and climate change modeling?

Arumugam: If you look at population, 90% live in cities, the percentage depends on how you define the urban setting, but that’s where most of the population is. The rest live in a rural setting. Rural setting naturally is also important, -- Most of the population is around the coast, that is the reason it is important to have a multi-plex. What we mean by multi-plex is not just stopping the flood and walking away. We need to take that perspective into protecting infrastructure. For example, the substation that supplies power to a city, or the water distribution system, and how it will impact the infrastructure. The multi-plex (are) ways to look at strategies, transportation pre-flood, during flood and post-flood. Together we can devise the evacuation maps and other things.

For example, we may have changes in flood frequency, but the changes may not be that significant in some locations, because there may be significant upstream storage.

WT: If the USGS drop more flow sensors into rivers and streams, is that important?

Arumugam: I think it is important. We rely on more data, not less data. We had a hurricane in 2018 in North Carolina which affected only a few locations. It’s not going to be manually possible to measure everything everywhere. More monitoring stations/sensors is going to be helpful, allowing the hydraulic models for flood protection and inundation.

WT: You are listed as a specialist, media contact for people that would want to know how cities and towns should get ready for an incoming hurricane. Is there a list of points that you could offer to these viewers, on how to get ready for something like a hurricane?

Arumugam: I can specifically talk in the context of North Carolina, for example. Depending on the hurricane intensity, you can expect certain flood-level outcomes.

Obviously, nobody is going to have a residential house in the 100-year flood plain, but on the other hand, if the 500-year flood plain had a small problem, would it become inundated? For example, when it continuously rains for a couple of days -- not a hurricane or anything—but continuous thunderstorms coming in can inundate the area.

The way we try to help is more along the lines of evacuation strategies and other things, looking at the flood forecast. Particularly at the state level, it involves looking at all the information available, how we can evacuate, plan the routing, without congestion.

WT: You were working on something called the Urban Flooding Open Knowledge Network, did it get built? Is it up? What was the premise for Urban Flooding Open Knowledge?

Arumugam: The idea is based pn the multi-plex, basically, we try to map the various characteristics of the area. Every different location has some problem with inundation, we try to quantify based on the flood forecast and flood frequency for the long term.

If the information is available, then based on the national water model forecast, then you can use that discharge, and flood stage information (for planning), to know how much flood you are expecting. That is the knowledge part. The application part relates this flood information to protect the area’s infrastructure, including evacuation planning.

I mentioned the national scale flood forecasting model, we work every day to get it updated.

WT: One of the premises of our media group is we gather information on water, and groundwater and we don’t have a paywall. Have you found in this industry any groups trying to monopolize the information, and then charge people for it? Is that becoming a thing in this industry?

Arumugam: That’s a good question. I don’t see that in this industry. On the water data front, most is collected by the agencies, but there are other agencies, and companies looking at (the data) and they have modeling software available, but if you want to get a model for the information, it's going to be proprietary. To some extent, there are models that are open source. In general, you don’t see much out there proprietary in terms of the data, holding on to the data.

There is free software, you can put in your location and get some flood map/info.

WT: To close, what is your advice for first-year students looking at urban flooding, these flood maps, thinking this is something they would like to get into, what is your advice to students, on how they can get to the level where they can make a difference?

Arumugam: I would say to someone entering research or graduate school, try to get a broad, interdisciplinary perspective. Have your specific course but look at the problem from an interdisciplinary perspective.

WT: Thanks for doing this Professor, I really appreciate this. Best of luck in your endeavors, everyone is going to need it.