Amid heavy flooding, Houston’s Texas Medical Center showcases what can be done with current technology.
The 15-plus inches of rain that overwhelmed Houston on Sunday and Monday and killed eight people showed just how vital early flood warnings can be.
Throughout the U.S., the National Weather Service issues flood advisories at the county and city level when it’s clear that a lot of rain is going to fall there. These big-picture announcements, though, are generally too broad and difficult for individuals to act on, says Philip Bedient, an engineering professor at Rice University and expert in flood warning systems. He’s working to fix that using a combination of cameras, radar, rain gauges, and digital modeling, and it’s already saved the massive Texas Medical Center a lot of rain damage. Broader implementation could reduce flooding deaths and losses elsewhere in Houston, and around the world.
Topography plays a huge role in determining which buildings are safe and which get submerged, so a truly useful warning system has to be more granular than county-level. Imagine if Google Maps only warned you of the presence of traffic in your county—that can’t help you decide what road to take.
“During these recent floods in Houston, we got these very general alerts on our iPhones saying the county is under a flood watch, but the county is 1,700 square miles,” Bedient says. “By the time you realize the water’s in your roadway and it’s six inches deep, it’s probably too late.”
Cities can achieve a higher level of detail with readily accessible technology. Bedient’s defense network for the medical center has been operating since 1998 and guided them through 50 major storms, he says. The system calibrates weather radar with data from water gauges and then uses a computer model to predict how fast the water is going to rise.
None of this matters unless the data lead to an actionable recommendation. Most people have little understanding of how the rainfall and flow rate translate into real-world forces; at what point does a wet nuisance turn into a current that can sweep your car off the road? Bedient’s platform condenses all the complex hydrological observations into a user-friendly traffic light system. All you need to know is that red means it’s not safe to go. For the medical center, that means they initiate a lock-down to keep the waters out.
Houston has plenty of other low-lying areas that could benefit from this type of warning system—Bedient has worked on systems for Sugar Land and Austin already. It makes sense to start with high-value vulnerabilities like hospitals and work out from there. The danger, of course is that poor populations don’t get access to this life-saving technology. That’s all the more possible because, as my colleague Kriston Capps wrote last year, Texas has minimal central flood control planning, despite having the second highest bill for flood claims in the 50 states. The lack of state-level funding and strategy leaves it up to cities to figure out for themselves.
The question, then, is why more cities aren’t clamoring for systems like the one at Texas Medical Center. Bedient says he can get it up and running for just $100,000.
“There’s no reason we couldn’t do it on every bayou in town,” says environmental law professor James Blackburn, who co-directs Rice’s Severe Storm Prediction, Education, and Evacuation from Disasters Center with Bedient. “It should be absolutely transferable to all major urban areas in the United States.”
There are some technical constraints, though. The sensors need power, so they need to be close to the grid or have batteries. And the political barriers are more salient.
For starters, an automated monitoring system still requires human labor to keep it maintained and powered, says electrical and computer engineering professor Elizabeth Basha, of the University of the Pacific. As a doctoral student at MIT she worked on building an automated flood sensing system in rural northeastern Honduras (that project halted when the local partner lost interest). Like any equipment, if it’s sitting alone in nature for months on end it will see wear and tear. Regular inspections are needed to be sure the tools are up and running in event of an actual flood. That requires long-term budget allocations, just like any other infrastructure investment.
Persuading governments to adopt new technology can be hard, but it’s even harder when that technology means the difference between life or death. “There’s a high cost to being wrong,” Basha says, so leaders need to be sure that the new system would work as well or better than the existing, less granular approach. There’s reason to think it will, but we don’t have a ton of data yet.
The best approach may be for more large institutions like the Texas Medical Center to invest in this technology for their neighborhoods, racking up more of a track record for cities to look at. American cities already have considerably more flood warning resources than their counterparts in the developing world, but they aren’t as far along as they could be. As a changing climate makes deluges like Houston’s more common, city residents may soon start demanding more.