Evacuate or Stay?
How the most important question ahead of a hurricane is killing us
Shortly after the first attack on the World Trade Center complex in 1993, the Port Authority of New York spent over $100 million beefing up security in and around its buildings. To protect against another vehicle bombing, giant five-ton, steel-reinforced planters were installed in a daisy chain enclosing their perimeter. A new custom movable gate permitted emergency-only vehicle access through a wall of heavy video surveillance. A repeater system was put in place to boost radio signals of first-responders entering the buildings. The investment in new technology would stave off, if not outright prevent, another attack. At least that was the thinking. Until the unthinkable.
The impact of the planes on September 11th, 2001, severely impaired much of the new technology and the confusion that followed rendered the all-important repeaters inoperable. As Amanda Ripley writes in her book The Unthinkable, on 9/11 most civilians in the twin towers had no idea how to evacuate. Thousands couldn’t even find the stairwells. Of those that did, many wasted precious time evacuating upward instead of taking the clearest route down. It was a veritable ant hill of confusion. Alan Reiss, the man who oversaw operations for both towers on September 11th, put it plainly in his testimony to the 9/11 Commission: though technology and equipment had improved since 1993, evacuation protocols hadn’t. Fire drills were held twice a year, but according to a 2012 Columbia University study, 94% had never exited the building as part of a drill.
There’s a common thread among disasters. A cheap strip of lights guiding you to the nearest exit can go a long way in increasing your survival odds during a fire. Reading the safety information card and locating your nearest exit before takeoff can significantly increase your odds of surviving a plane crash (yes, plenty of people survive airline accidents, about 95% according to a comprehensive NTSB study). In all disasters, knowing when, where, and how to escape is key, which brings us to the hurricane problem.
Hurricanes are survivable events. They don’t strike out of nowhere. We have a battalion of satellites, buoys, and airplanes to track their every move from their infancy over the sea to their kamikaze-like demise along our coastlines. Forecasters man 24-hour weather offices that, in most cases, give days of warning of an impending storm. Yet some 2,000 people have died, either directly or indirectly, from hurricanes in modern, 21st century-America.
It’s easy at first glance to write off the flooding of New Orleans after Hurricane Katrina. “It was a man-made disaster,” is what I hear most often (in reference to the catastrophic levee failure and subsequent response). Perhaps it was. I won’t wade into those political waters. But two days before landfall, when Katrina topped the scale as a Category 5 hurricane, emergency planners were expecting the worst Mother Nature could throw at them. They knew at least some floodwater would seep into the city . It’s why the city issued a mandatory evacuation for all within its borders. By and large, the people of New Orleans listened. Over 80% of the city filed out ahead of Katrina. That level of evacuation compliance is almost unheard of. In most instances less than 50% ordered to evacuate before a hurricane actually do so. During Sandy, fewer than one-in-three New Yorkers heeded mandatory evacuation orders.
While it’s important to know when to evacuate, it’s equally important to know when not to evacuate. Take Hurricane Rita back in 2005. It was the Category 5 monster that blew through the Gulf of Mexico less than a month after Katrina ravaged the Gulf coast. The pictures of bodies floating in city streets were still salient on the minds of most Americans, especially of those under threat of the next big hurricane. When the city of Houston, the fourth most populous city in the country, said the time for waiting was over, people left. In fact, too many people left. What ensued was the largest evacuation in United States history.
But in the case of Rita, nearly half of those that evacuated (an estimated 1.5 million of 3 million total), as many as evacuated in total for Katrina, were not actually ordered to evacuate. These “shadow evacuees” put a strain on the roadway infrastructure that, even in the event of a landfalling hurricane, was not built to handle the evacuation volume. Evacuees were planted in parking lots of highway traffic for up to 36 hours trying to flee inland. Vehicles broke down or ran out of gas atop hot asphalt pavement under the scorching, September Texas sun. The result: over 100 evacuation-related fatalities.
Hurricanes are unique events. They have the capacity of shutting down major American cities and causing severe damage to the nation’s economy. They have wreaked more damage to this country over the past 30 years than nearly all other weather phenomena combined (over half a trillion dollars since 1980), including drought and severe storms. They’re responsible for four of the ten deadliest disasters, natural or otherwise, in U.S. history. Yet despite the fatalities, they’re the only weather phenomenon that we routinely evacuate in advance of.
As meteorologists, we speak frequently and passionately about improving our forecasts. The U.S. should invest more in its weather models. Five day hurricane forecasts should be extended to seven days. Here’s the reality: the tragedy of Katrina wasn’t caused by a bad forecast. The nightmare during Rita wasn’t brought on by a faulty weather model with an outdated convective scheme. We can make a perfect forecast and people will still die. We have our weather heads buried deep in the forecast sand. A forecast is a process, not an outcome. Saving lives is a team effort and our teammates desperately need us.
During my time with the state of Florida Emergency Operations Center in the mid-2000s, I learned quickly that evacuations are a tricky business. Why people stay or why they leave ahead of a hurricane is a complex social and psychological web. Social scientists have spent decades studying the problem, continually updating their social algorithms to untangle the messy web. Predicting the movement of hurricanes that follow the laws of physics is a cakewalk compared to the largely unpredictable realm of human behavior.
It’s a vexing problem. For starters, most don’t realize that we primarily evacuate because of a hurricane’s storm surge (ocean water pushed over normally dry land), not its winds. So determining who must go means first determining where the water might go. That’s the physical science side of evacuations. Then there’s the engineering side. What public buildings will make viable shelters and how will roadways handle the mass exodus? Then of course there’s the human side. How many will leave and how many will stay? What will increase the odds of someone leaving that should and decrease the odds of someone leaving that shouldn’t?
Answering these questions requires real investment. As with weather forecasts, it takes money to draw up good evacuation plans. FEMA is alloted $2 million a year through its National Hurricane Program to handle all hurricane evacuation assistance, including hurricane evacuation plans, to State, Local, and tribal governments. Any changes to your local evacuation zones and evacuation procedures probably come from this small pot of money. The development of decision support tools and software used by emergency mangers nationwide comes from here. It includes not only social science and engineering research, but all storm surge modeling performed by the National Hurricane Center. This includes updates to the same storm surge models used to make government forecasts during a hurricane event.
Consider this. While the budget for hurricane research and forecasting has over doubled (roughly $15 million to $35 million) since Hurricane Katrina, the budget for improving evacuations has flatlined since 2005. As a result, updates to regional evacuation studies have been backlogged. Southeast Louisiana, which includes the Greater New Orleans metro area, hasn’t updated its Hurricane Evacuation Study since the turn of the century. That means the $14 billion of improvements to the levees since Katrina aren’t considered in their current hurricane evacuation zones. That’s like changing the design on all Boeing 747s but waiting 15 years to update the safety information card. Sorry, the nearest exit is really behind you. This isn’t a little problem. This is a big problem.
But it’s not a FEMA problem. On the contrary, FEMA has been working overtime to crank out updates and modernize its technology to increase efficiency. They’ve sought out (and received) support from the U.S. Army Corps, the National Hurricane Center, and the states they support. They’re doing their best with the resources they have. But they need more resources. Hurricane plans don’t write themselves, especially good ones.
It’s easy to write a plan for the college educated, those who own a vehicle, have insurance, and have enough squirreled away to afford a three, four, or five night out of town stay for their family. It’s not so easy to write plans for the most economically disadvantaged or those with disabilities — in other words, for what’s real. The reality is one-in-ten Americans don’t drive and 60% of American households own a pet. So when we say leave, what’s the plan for those who can’t drive or who own pets?
Government alone isn’t the answer. It’s why the entire hurricane community, including the private sector, non-profits, and local communities, spend an inordinate amount of time during the “off season” preaching preparedness. With hurricanes, we truly are our brother’s keepers. Even so, government serves its purpose. Unlike fire drills we can’t practice large scale hurricane evacuations but we can ensure our plans are up to code, based on the latest design, and communicated in advance to those in harm’s way. The outside threat may not change but our understanding of it does.
Cities change; levees are updated, buildings codes are strengthened, lanes are added, populations are shifted. Changing the outcome means changing behavior. All of the technology in our palms, pockets, and purses won’t do us a lick of good if during the fire we can’t find the strip of lights leading us to safety.
Katrina wasn’t the unthinkable but the unthinkable will come. The question is are we willing to change the outcome before it does?
