Tweets and posts
Measuring social media’s effectiveness during and after the flood
In their personal experiences in Columbia during last October’s flooding, the Arnold School of Public Health’s Heather Brandt and Brie Turner-McGrievy were struck by how well many individuals worked together in responding to the disaster.
“When you woke up on Sunday morning during the flooding here, people were in the boats that they had unhooked from their own trailers, and were off rescuing people. These were citizens doing that,” Brandt says. “Within 24 hours, huge teams of self-organized neighbors had gone into flooded homes and said, ‘We have to get this drywall out of your house because mold sets in within a very short period of time.’”
Five days of incessant downpours had struck the Palmetto State hard, and South Carolinians responded with months of highly orchestrated rescue, repair and recovery operations, with an almost preternatural collective sense for finding those who needed help most.
That level of self-organization, Brandt and Turner-McGrievy hypothesized, probably had some roots in social media, a tool employed by emergency responders, government agencies, nonprofit groups, media and everyday citizens during the storm and its aftermath. But when they organized a team to document and analyze the social media response to the flooding, they found that, like the flood itself, its volume was overwhelming.
“The sheer number of tweets and Facebook posts would break a computer,” Brandt says.
So they scaled back their approach, but even when they focused solely on the Twitter platform, restricted the scope to Midlands tweets, considered tweets only from a limited number of hashtags and included just one in every 10 in the analysis, the team still had thousands more tweets to consider.
The heft of that dataset, representing just a sliver of all of the social media response, underscored the amount of engagement that online platforms supported during the crisis. In the preliminary stages of coding the data, the team recognized just how much value social media brought to many facets of the recovery.
The sheer number of tweets and Facebook posts would break a computer.
Social media was a key part of informing the community about the forecasts concerning the approach and ongoing severity of the rainfall, which were remarkable in their accuracy, they found. Social media was integral in organizing volunteering, FEMA assistance, resource distribution, cleanup and fundraising. It also played an important part in helping people understand road conditions and avoid danger while traveling.
As Brandt and Turner-McGrievy work through the data, they’re identifying organizations that exhibited highly effective use of social media during the flooding. One of the most prominent is the S.C. Emergency Management Division, which not only spread valuable information as it was needed, but also helped refute misinformation (such as the erroneous notion that the Lake Murray dam was at risk) as it arose on social media channels.
Looking at the disaster through a public health lens, the team is working to develop a guide to best practices along with documenting the social media response that accompanied last year’s disaster. An avid Twitter aficionado herself, Brandt (@BlondeScientist) wants the success she saw firsthand to propagate even further.
“My husband and I worked with the My Carolina Alumni Association during the recovery, and we used Facebook and Twitter to guide donations and deliveries,” Brandt says. “We would find somebody who said, ‘We need water, or we need this here.’ And I would reply and say we would be happy to help you, can you direct message me a street address? And then we’d send a truck off with the supplies.
“That wouldn’t have happened maybe even five years ago. But it’s definitely happening now, so let’s figure out how we can use it most effectively.”
Closing tomorrow’s floodgates
Destruction from the flooding was exacerbated by numerous dam and levee breaches in Lexington and Richland counties. Hanif Chaudhry, a civil engineering professor and associate dean of the College of Engineering and Computing, is leading a team of researchers who aim to learn as much as they can from the dams that failed. The researchers didn’t have a moment to spare once the rain stopped falling.
“It was very time-sensitive,” Chaudhry says. “When there is a failure, if they want to rebuild, they will start right away with construction. Or if the dam is gone, the flow of water will change the characteristics of the site. Or if some of the structure is left, it might present a hazard that the owner will move in quickly to remove.”
Fortunately, Chaudhry and his engineering colleagues were already in the midst of two similar studies of dam breaches, including a $3 million project funded by the National Science Foundation. They had the manpower, training and laboratory equipment to quickly collect and analyze field samples.
As part of the S.C. Floods Initiative funded by the university’s Office of the Vice President for Research, Chaudhry’s team visited 14 sites with major damage to earthen levees, nine of which involved total failures. Data collection included dimensional measurements and samples of the materials present in the remnants of the structures.
Some of these dams were in series. So even if a given dam is in good shape, if an upper dam fails, it comes down as a major wave, and if that cannot be handled it overtops, and then the next one. So it can become a cascade. The dams are a system that we need to properly manage.
The full analysis is still in progress, but the consequence of overwhelming the capacity of a dam’s spillway to move excessive inflow to the downstream side of the embankment was clear. The result, Chaudhry says, was “overtopping. That’s why most of the dams failed.”
He hopes that a complete analysis of the dam materials and the failure conditions might lead to design recommendations that will mitigate damage in future flooding events. Moreover, hydrological modeling of levee capacities at the point of failure should provide leaders with the capacity to better manage dams and levees that are highly connected.
“Some of these dams were in series,” Chaudhry says. “So even if a given dam is in good shape, if an upper dam fails, it comes down as a major wave, and if that cannot be handled it overtops, and then the next one. So it can become a cascade. The dams are a system that we need to properly manage.”
Stages of loss and resilience
Peter Duffy didn’t think words on paper could adequately convey the personal stories of loss, struggle and resilience in the face of last fall’s flooding.
That’s why the USC theatre professor assembled a creative team to document the human suffering that followed the flood.
“It’s one thing to know that 11 trillion gallons of water fell, and that it’s enough water to have quenched the drought in California,” Duffy says. “But it’s another thing to hear what the flood has meant for people, and for so many people it’s still going on. You can’t really convey that information to the public without mediating it.”
Duffy led a team that interviewed three dozen Columbia residents hit hard by the disaster. The goal is to use art to express those experiences to the rest of the community, with three performances planned for a one-year commemoration of the flood in early October.
Working through transcribed and coded interviews, the team identified recurring themes, and Duffy created composite characters who will take the stage in the upcoming performances. Dance and photography will be part of the presentations.
Listening to people tell their stories revealed challenges the researchers never anticipated.
“Some people had family come from out of state to help out at the drop of a hat, but for others that wasn’t the case,” Duffy says. “When you think about devastation of the flood, you’re thinking about shoveling out mud, not necessarily another layer of heartbreak on top of all that: having to deal with your family not being disinterested but they’re just not here.”
Resilience is a recurring theme in the interviews, Duffy says.
“That’s been an important question: ‘What has gotten you through this?’” he says. “And what does the city still need to know?”
It won’t just be the city of Columbia that will get those answers. The team is working with C&T, an England-based theatre company that is providing an online platform to create a map highlighting similar projects around the world.
“We’re going to have a video of our play, which will be pinned to South Carolina,” Duffy says. “There are lots of similar projects happening around the world, looking at the impacts of weather-related phenomena or climate change related phenomena.
“Each performance will get pinned, so we’re creating a global interactive map that uses arts-based methodologies to share what’s happening globally.”
Ebb and flow
The flood as economic stimulus
First came the rain, then — in one tumultuous weekend — the flooding. But the economic effects of last fall’s weather event lingered long after.
Economists Doug Woodward and Joey Von Nessen at the Darla Moore School of Business are using data from the S.C. Department of Insurance and FEMA to better understand how the flooding has affected the state’s economy.
Those organizations track money flowing into the state to cover losses in agriculture, infrastructure and real estate, and the researchers have seen, in their preliminary data, that significant funding has arrived already. In terms of how those dollars might still affect South Carolina’s economy, though, it’s a question of when they will be spent. And there are plenty of places to spend.
“We see most of the damage in the Midlands and in Charleston, but approximately half of all of South Carolina’s 46 counties had some sort of damage associated with the flood,” Von Nessen says. “We are going to see unambiguous losses in wealth as a result because there were a lot of uninsured individuals who were affected that are not being reimbursed for any damages. In many cases property is going unrepaired altogether.”
The overall reduction in net wealth should, however, be accompanied by a stimulus to the economy, the economists believe, at least in the short run. After Hurricane Hugo in 1989, for example, economic activity ramped up, Von Nessen says, and he and Woodward are seeing some of the same in the early data from the recent flooding.
“Usually in natural disasters like these, you see a stimulus in two sectors. One is construction because damaged infrastructure is repaired and firms are brought in to do that work,” he says. “But we also are tentatively seeing a boost in retail activity as well, and a lot of that is housing-related goods and services.”
It is a stimulus, but I think it would also be appropriate to think of it as almost a mini-bubble. Many of the repairs are to damaged property that in some cases needed renovations anyway, repairs that people may have been looking to make down the road.So in a sense we see a borrowed future demand: you see a bump up, a short-term stimulus, and then a bump down, below the long-term averages, later on.
—Joey Von Nessen
One specific example of the stimulus that they expect to see in the final analysis is an increase in hiring.
Looking at employment going forward, we’re expecting about 2.7 percent employment
growth — our current rate now — to persist for the remainder of 2016 and into 2017,
and our anticipation is that the stimulus from the flood may have as much as about a quarter to a half of a percentage point bump on that baseline employment growth,” Von Nessen says. “So we might expect perhaps 3 percent employment growth or slightly more.”
That kind of economic bonus is nice in the short term, but is expected to be followed by an eventual downside, Von Nessen adds.
“It is a stimulus, but I think it would also be appropriate to think of it as almost a mini-bubble,” he says. “Many of the repairs are to damaged property that in some cases needed renovations anyway, repairs that people may have been looking to make down the road.
“So in a sense we see a borrowed future demand: you see a bump up, a short-term stimulus, and then a bump down, below the long-term averages, later on.”
Weathering the storm
The North Inlet, which connects the Atlantic Ocean to the combined river output of the Pee Dee River basin in Winyah Bay, is an estuary that has remained essentially untouched by commercial human activity. Located just east of Georgetown, S.C., its salt marsh has been the subject of meticulously detailed study since 1981 by Baruch Institute scientists, who think that the marine ecosystem has existed largely as it is now for hundreds, and perhaps thousands, of years. That combination — a lack of human influence and a meticulously detailed, 35-year record of scientific study that has shown evidence of small perturbations more recently — is almost unmatched globally, making the North Inlet one of the world’s most important sentinels for climate change study.
One of the most scientifically important marine ecosystems in the world got a desalinizing wallop from last fall’s deluge: The creeks of the North Inlet of Winyah Bay had low-tide salt concentrations drop to levels never before seen in 34 years of measurement.
Since then, researchers at the Belle W. Baruch Institute for Marine and Coastal Sciences have stepped up their monitoring of the North Inlet’s salt marsh, looking for signs that the freshwater inundation might have fundamentally changed the estuary.
We saw a number of species of fish and at least one shrimp that were records for North Inlet. They had never been seen there before.
With funding from the S.C. Flood Initiative, the Baruch Marine Field Laboratory has augmented existing monitoring stations and collected samples more frequently to assess population changes in shrimp, crabs, fish, plankton and microbes dwelling in the estuary. Baruch researchers found that, in the days and weeks soon after the flooding, the altered estuary provided ideal conditions for some new visitors.
“We saw a number of species of fish and at least one shrimp that were records for North Inlet. They had never been seen there before,” says Dennis Allen, co-director of the field laboratory. “They include channel catfish, white perch, white catfish, juvenile herrings and shads — organisms that are much more typical of the rivers.”
The low-salinity waters at low tide that were hospitable to those more freshwater-oriented newcomers did return to normal saltiness by mid-to-late October 2015, but the effects of the deluge were far from over, Allen says. A wet El Niño winter followed, and several storm systems — though each released considerably less precipitation than the original flooding event — caused salt levels to repeatedly crash, with extremely low readings still being recorded into March 2016, some seven months after the initial deluge.
“Normally there wouldn’t have been much of a signal in terms of salinity depression,” Allen says, “but because of the October flood the soils throughout the watershed were so saturated that almost anything that hit the ground ran off.”
By April, salinity levels had returned to what scientists deemed normal, and they turned their focus to the question of whether the flooding event might have caused long-term changes in the estuary. Early returns indicate otherwise.
“Pending the full analysis, I think what we’ve seen here was a short and intense perturbation, but one that was not enough to reset the system at a different level of organization or function,” Allen says. “So it was a shock to the system, but it appears to have recovered, and at least is on its way to returning to what we recognize as typical over the past 35 years.
“I guess the good news for North Inlet is, it weathered the storm.”
The S.C. floodwaters in October 2015 took away life and property, but they also claimed something you can’t see as readily: peace of mind.
Looking to assess impacts on some of the most psychologically vulnerable victims of the disaster, College of Education faculty member Jonathan Ohrt is leading a team of researchers focused on the Richland and Lexington county school systems.
“In a natural disaster, schools tend to be a meeting place where there are resources for students and their families as well,” Ohrt says. “It’s just a place that people from the community have come to rely on.”
Interviewing mental health professionals who were primarily school counselors, Ohrt’s team is documenting some of the invisible long-term wounds that the flooding imprinted on young psyches.
“One student, whenever it rains, now gets very nervous because she’s wondering if something bad is going to happen. ‘Are we going to have to leave our house?’” Ohrt says. “Another counselor is working with a family that still isn’t in a stable place. They’re having to live in hotels — eight months after the event. Most of us, I wouldn’t say we’ve forgotten, but we’ve moved on in many ways.”
Members of Ohrt’s team, which included Department of Educational Studies colleagues Dodie Limberg and Ryan Carlson, can readily empathize with students and understand the school professionals helping them work through the situation, as well. Two team members were school counselors and another was a mental health counselor in K-12 systems before moving into academia. Coincidentally, all three were in central Florida (in different locations) in 2005 and experienced Hurricane Wilma as it plowed across the state.
“So we were kind of on the front lines before, and now we’re looking at it from a
researcher’s perspective as well,” Ohrt says. “We all had some personal experience
and actual work experience, which is one of the reasons we felt compelled to work