There is More to Floods than Meets the Eye
This year’s Climate and Society class is out in the field (or lab or office) completing a summer internship or thesis. They’ll be documenting their experiences one blog post at a time. Read on to see what they’re up to.
Sarah Toh, C+S ’17
The best maps tell stories. They take us on journeys in places both familiar and strange, revealing information that might not have been evident before. They also tell these stories remarkably quickly, allowing us to gain a better understanding of situations or make decisions in short periods of time. Not all types of information lend themselves to being mapped, but fortunately for me, the information that I dealt with during my internship, did.
In the aftermath of natural disasters, maps tell a variety of stories from ones about the people and places who need relief to the causes of disasters to ways to reduce their impact in the future. The project that I have been undertaking at the International Research Institute for Climate and Society (IRI) falls into the last three categories. Using a variety of geospatial information — data referenced to specific places — I attempted to create a map to improve our understanding of the factors that influence the impacts of flash floods.
The idea seemed simple: using case studies of past flash floods, I would tell this story in Google Earth Engine. Based on a review of a variety of events, I decided to create a story around the floods that took place in Tennessee in May 2010. This was based on considerations of the availability of information, severity of impacts and the interaction of different types of floods.
However, it is difficult to tell a story if the main character is not well-defined. Flash floods are defined differently in various parts of the world and they can also be difficult to separate from their comrades, riverine floods and urban floods, both in theory and in practice.
One of my first steps involved defining the various flood characters implicated in this story. I did this using historical flood warnings, news reports and an understanding of the different characteristics of floods.
Next, I had to set the scene for the story. Geospatial data help to do this in any map by providing background information. However, gathering this data can be challenging, both in terms of availability and compatibility. There’s a wealth of data in more developed parts of the world, and in places that are better prepared for disasters. This data is not always specifically designed for studying flood events, however and may need to be manipulated to suit this purpose.
In the case of the floods in Tennessee, availability was not a challenge. However, reconciling disparities between different datasets, such as conflicting reports about flash flood and flood events, required more research, and ultimately some narrative license in the form of informed decisions. Choosing what geophysical and socio-economic data to include — and exclude — was also a crucial step, as it would influence the kind of story that the map would tell.
Making those choices required an editorial eye to ensure that the story did not overwhelm any viewer with too many sub-plots, while retaining enough information to be meaningful. At the same time, I had to be careful not to introduce too much bias by cherry-picking the types of information I included, a move that would tell a misleading story.
For example, including geophysical factors such as elevation, streamflow and precipitation, along with data about land surfaces and infrastructure, would allow for a more comprehensive analysis. But too much of that information could end up telling a story about Tennessee’s landscape, rather than one about a specific flood event and the factors that influenced its impacts. The details of these problems, when considered on their own, can seem trivial from the confines of a desk. However, they also gave me a newfound appreciation for the work that goes into maps that many people only look at for a few seconds.
It is tempting to think that more data will lead to better conclusions, but one thing that has emerged over the course of my internship is that it does not always produce any conclusions. And when it does, those conclusions can be quite different from initial expectations. In this case, qualitative comparisons did not produce any correlation between poverty levels and the severity of impacts, challenging some initial assumptions. The map also gave rise to more questions than it answered, such as how to differentiate between flood and flash flood impacts, and whether the distinction matters.
I started off this internship hoping to gain some familiarity with mapping software. As the process has expanded, it also revealed that part of the skill in creating maps lies in choosing the data that underlies each chapter of the story, and that foundations such as strong characters are just as important. The map that I am working on is not as beautiful or as informative as many others I’ve seen, but it is a just one book in a growing library that will help us understand the most serious weather hazard in terms of deaths in the U.S.