Warning Signs: Technology speeds disaster alerts, response
Published 8:27 pm Monday, May 13, 2013
Caitria O’Neill remembers her reaction to hearing tornado warnings on June 1, 2011. She went to the grocery store, she said, “because I live in Massachusetts, and we don’t get tornadoes.”
By the time she got home, hail the size of baseballs was falling. Even as she watched a lamppost blow past, what was happening didn’t register.
Fortunately, her father got her inside and into the basement before the tornado blew out the windows, ripped off the roof and knocked their house slightly off its foundation.
O’Neill, then 22, and her family emerged to devastation. Her town of 8,500 people, Monson, Mass., had just withstood an EF-3 tornado that destroyed more than 60 buildings including town hall, damaged 200 others, and would lead to one death.
Pressed into service, O’Neill helped organize volunteers gathering at a church across the street. What she observed about the process inspired her and her sister Morgan, an MIT graduate student, to start an online software company, recovers.org, to help people prepare for disasters and respond more effectively.
In a world that can be app-addled, it’s tempting to dismiss the idea that technology will solve all problems. Yet, technology has changed the way Americans get ready for disasters and respond to them – with more precise forecasts, personalized weather warnings and more efficient recovery efforts. And it will continue to help us be more prepared.
Recovers.org is a fledgling company with a web-based software that helps individuals and communities plan for emergencies, then after a disaster matches victims’ needs with volunteers and donations. Its technology can’t compare to major breakthroughs in forecasting that occurred in the 1990s, like Doppler radar, which lets us “see” tornadoes before they form. Or the Geophysical Fluid Dynamics Laboratory model, the first computer simulation to successfully merge information about the ocean and atmosphere to predict hurricane courses and intensities.
But Brian M. Brooks calls the web software “a godsend.” The city manager in Forney, Texas, Brooks used recovers.org when a tornado outbreak destroyed about 20 houses, a business and a school in his East Texas community on April 3, 2012. Coordinating with Community Life Church, the city set up a registry of people who needed help and people willing to volunteer goods or services. Cooper Taylor, the church’s mission director, said the website greatly reduced unwanted donations and storage problems.
Recovers.org is one example of technology pushing preparedness and recovery into the hands of communities and individuals. Even small towns now use established databases and GPS mapping tools to do things like track private storm shelters. Communities can respond to disasters differently, too. The same GPS technology can be used to plot downed trees. At least one company sells a system that tags trees and other debris with bar codes to ensure haulers don’t overcharge local governments and FEMA.
Specific technologies don’t always succeed, of course. “The challenge with any software is the plan is only as good as people’s buy-in and people’s knowledge of how to implement it,” said Andrew Sachs, vice president of government services at Witt O’Brien’s, a consulting firm that specializes in public safety and crisis management.
But some technology only needs to spread to be effective. That’s certainly true for storm warnings, which now reach individual pockets and purses.
Since April 2012, the National Weather Service has been able to send alerts about weather emergencies to people with newer smart phones. (About 55 percent of Americans have smartphones, according to ComScore.) Messages can also be sent for local emergencies, AMBER alerts and presidential alerts during national emergencies.
Smartphone alerts are geographically tailored. The weather service also has devised a “polygon” warning system, tied to cell towers, to make warnings for things like tornados, floods and severe weather more specific than the old, county-based system.
Private companies also offer telephone-driven warning systems. For example, TechRadium’s Immediate Response Information System makes automated calls when the weather service issues alerts. In Royce City, Texas, officials are using TechRadium’s technology instead of warning sirens, which the city’s fire chief and emergency management coordinator, Richard Bell, said were more expensive and only worked for people who are outdoors.
The weather service is experimenting with social media, as well, to reach people who don’t watch weather on TV or own weather radios. In Oklahoma – where the weather service has 32,000 Facebook friends and the Twitter handle @nwsnorman has 13,500 followers – a test in March found that a Facebook message reached 200,000 people in just a few minutes.
Even those hyper-personalized alerts will not mean the end of long-time staples of weather warnings like radios and sirens. In part, that’s because cell phone networks can be overwhelmed by traffic or knocked offline in a crisis, said Kim Klockow, a PhD student in geography at the University of Oklahoma who is researching how people respond to tornado warnings. Klockow’s work after the 2011 tornadoes in Alabama and Joplin, Mo., confirmed earlier studies showing that people want multiple sources of information about storms. More than half of those she interviewed first learned of those tornadoes from sirens then checked other sources such as television news or their friends.
Nor has technology created major changes in the way storms are forecast, according to scientists, at least since the advent of Doppler radar and integrated models for hurricane forecasting. But a great deal of work has gone into making those models better.
The weather service is rolling out changes to Doppler radar called dual-polarization, which distinguishes between hail and rain. It’s also testing phase array radar, used by the U.S. Navy, to see if it improves tornado forecasting. Traditional radar involves a dish that spins, but phased array radar uses panels that measure atmospheric volumes in about one-fifth the time as current radar, said Harold Brooks, senior scientist at the weather service’s forecast research and development division in Norman, Okla.
Phase array radar might make it easier to forecast tornadoes a week or two in advance, as can happen with hurricanes, or even make it possible for communities to have tornado days, like wintry areas have snow days, said Brooks. Phase array also could consolidate multiple radar systems – such as those used for air traffic, defense and weather – into one.
Growth in computer power alone is improving hurricane forecasting, and models are improving in their ability to predict the paths of storms. Progress is slower at predicting intensity of the storms, which is driven by ocean temperatures. In particular, the models don’t do a good job predicting whether hurricanes will go through rapid increases or decreases in force.
That uncertainty can weaken public attentiveness and ultimately cost money, said George R Halliwell Jr., a research oceanographer at the Atlantic Oceanographic and Meteorological Laboratory in Miami. Halliwell has developed a model, currently in testing, that he hopes will help address the issue.
The difference between a Category 1 and Category 3 hurricane is substantial. And while a Category 5 hurricane in the Gulf Coast should trigger evacuations, if such a storm degrades to a Category 1 a few hours later, said Halliwell, “people just wasted a lot of money” evacuating.
Even with advances in technology, forecasters, scientists and emergency officials still face a dual challenge: When storms and disasters loom, they must convince a sometimes skeptical public of possible danger. Even before that, they must coax people – like Caitria O’Neill before the tornado hit her town in Massachusetts – to prepare for the unthinkable.