Frustrated tweets led scientists to believe that tidal floods along the East Coast and Gulf Coast of the US are more annoying than official tide gauges suggest. Half a million geotagged tweets showed researchers that people were talking about disruptively high waters even when government gauges hadn’t recorded tide levels high enough to be considered a flood.
Capturing these reactions on social media can help authorities better understand and address the more subtle, insidious ways that climate change is playing out in peoples’ daily lives. Coastal flooding is becoming a bigger problem as sea levels rise, but a study published recently in the journal Nature Communications suggests that officials aren’t doing a great job of recording that.
The Verge spoke with Frances Moore, lead author of the new study and a professor at the University of California, Davis. This isn’t the first time that she’s turned to Twitter for her climate research. Her previous research also found that people tend to stop reacting to unusual weather after dealing with it for a while — sometimes in as little as two years. Similar data from Twitter has been used to study how people coped with earthquakes and hurricanes.
This interview has been lightly edited for length and clarity.
Why were you interested in studying coastal flooding?
We’re trying to get at some of the negative effects of climate change that are widespread but not necessarily disastrous. They’re having these negative consequences across large populations, they’re interrupting people’s daily lives, they’re annoying, they’re causing some damage — maybe not a huge amount. Those types of impacts are pretty important, even though any individual one [is not necessarily] particularly dramatic. That type of pervasive and chronic nuisance flooding that is just getting worse and worse and worse over time as people are trying to go about their daily lives is something that we’re interested in understanding.
Why did you turn to Twitter?
It’s an interesting source of data to complement the more standard measures that come from tide gauges that [the National Oceanic and Atmospheric Administration] puts up to measure tide heights. If you have enough data, then you can kind of get a very fine resolution of where this flooding is happening. Whereas the tide gauges, you might have maybe one in a county typically.
The main reason I like [Twitter] as a source of data is that it integrates not just a measure of typical exposure — which is “Did the water come onto the land in a place where it wasn’t supposed to be?” It combines that typical exposure with a measure of social consequences. If the water is where it shouldn’t be, but it’s in a place where there aren’t any people and there aren’t any roads, that’s really different from that same amount of water that’s in a busy road in the middle of rush hour in the middle of the city. Measuring “What are people noticing? What are people are talking about?” Twitter can kind of give us this aggregated measure of what those social consequences of that particular flood are.
Are there any limitations to using Twitter?
Not everyone in the US is on Twitter. It’s a subset of the population. The demographics of who is on Twitter versus the general population, we don’t have super good information about that. But it’s probably kind of skewed in some ways related to income, related to race, and related to age. So we might be getting a slight, somewhat bias. Generalizing our results to the entire population, you may want to be a little bit careful about that. I would also say that, which was actually pointed out to me by another reporter, is that we’re only looking at the English tweets for this study. And typically in Florida along the Gulf Coast, we might get a lot from tweets in Spanish. And that’s something I think we’ll think about going forward just to try and get more of the population represented in our study.
What do you hope might happen with your work? How might it impact communities?
For any particular community, I’m sure people who work in that community have a pretty good sense of where flooding is happening, under what conditions, and who is being affected and how. What we get here is being able to say something across larger geographic areas but still keeping some kind of local realism in the sense that what we’re doing is aggregating information from these local areas. It might be a way of tying these local experiences to a larger conversation at the national level. How do we measure flooding? Are we measuring it right? There definitely seem to be a few cases where the local experience is quite different from what the flood gauges, the flooding estimates show. How we measure this line of flooding at a national scale, those measures are not particularly good and we might want to be thinking better about how to do that in a standardized way.