15 September 2017

Getting evacuations right—in time for the next one

Jeff Terry

Jeff Terry

Jeff Terry is a professor of physics at the Illinois Institute of Technology, where his main research focus is on energy systems. His group works to develop new ways of dealing with radioactive...


Many scientists now study topics they never considered early in their careers, their new interests brought on by the changing climate. I, for example, never expected to review the literature on evacuating people from natural disasters. As a nuclear scientist, I started looking at evacuations to understand what could happen if a nuclear reactor were to have a catastrophic meltdown in the United States. I wanted to know who should be evacuated, from what area, and when, as well as where they should go and how to get them there. It turns out, though, that there aren’t that many examples of evacuations from areas around nuclear plants. (The Three Mile Island accident and the Chernobyl and Fukushima disasters are exceptions.) I expanded my search to include types of evacuations that occur more frequently, including from tsunamis, wildfires, and hurricanes.

Unfortunately, as the climate changes—and for this analysis, it doesn’t matter if you think observed climate change is anthropogenic or natural—I expect we’ll be able to gather more and more data in the coming years, as evacuations for wildfires, large hurricanes, floods, and epidemics become much more common. In the past month alone, North America has seen tens of evacuations due to flood, fire, storm, and earthquake, with many more in other parts of the world as massive floods have hit South Asia. We can, at least, use what we have learned to make the next evacuations go more safely and smoothly.

Reading studies of past disaster preparedness efforts, one of the first things that becomes obvious is that evacuations themselves are deadly. The evacuation following the 2011 Fukushima nuclear power plant disaster killed between 600 and 1600 people, depending on how you calculate. At least 400 Louisiana residents who fled from Hurricane Katrina in 2005 died within five weeks, according to reports from out-of-state coroners. Memories of Hurricane Rita, which hit Houston in 2005, were highly relevant to the debate over whether or not Houston should have been evacuated ahead of Hurricane Harvey late last month. In 2005, hours before Hurricane Rita hit Houston, approximately 2.5 million people took to the freeways. The death toll in this evacuation exceeded 100, and some people were trapped on the roadways for 24 hours.

When we compare these evacuation-related deaths to the average number of deaths expected from hurricanes in the United States—on average, 46 per year over the last 30 years—deciding when and how to evacuate becomes a very difficult decision. Globally, deaths due to natural disasters have decreased decade by decade for most of the last century. Better building codes have saved lives, and governments have identified targeted evacuation zones well in advance of storms. As things currently stand, Harris County, which encompasses Houston, reports 30 deaths related to Hurricane Harvey. Based on this figure, and the 100 who died trying to evacuate from Houston ahead of Hurricane Rita, it is hard to argue with Houston officials’ decision not evacuate this time around.

Of course, it would be better to bring the death toll from natural disasters down further. To do so, we must, as a society, improve the way we handle evacuations. Here are some ways to do that.

Help those who need it most. Natural disasters have a disproportionate effect on the elderly and others with little access to transportation. Disaster preparation must account for protecting those with special needs, who include people with disabilities and medical conditions, people who live in care facilities, the homeless, prisoners, and people who don’t have vehicles. We should not be willing to condemn whole groups of people to ride out disasters on their own because we lack the means to transport them to a safe location. As a 2009 US Department of Transportation Report pointed out, “During medium to large evacuations, every transportation system will be overwhelmed especially when evacuating special needs populations,” but though that was written eight years ago, cities haven’t necessarily heeded the warning. Governments must invest to ensure they have enough properly outfitted vehicles to move special-needs residents. Alternatively, if we decide that moving these populations is unwise due to their vulnerability, we have to develop hardened facilities with the equipment, including back-up power, to support special-needs residents for many days. Unfortunately, the failure to account for special-needs individuals in a nursing home in Florida resulted in eight deaths in the wake of Hurricane Irma.

The poor also need special consideration, as evacuations are not cheap. They require an initial outlay of funds for transportation, hotels, food, and other essentials while displaced from home. Since the poor tend not to have the kind of savings that could cover these expenses, they end up left behind in dangerous situations. College students, too, may lack the funds for an emergency move. Some colleges are starting to develop evacuation procedures for students without the means to return home. Will other institutions follow their lead?

Get out of the car. As a country, Americans love their cars, which is one reason “every transportation system will be overwhelmed” during evacuations. Freeways become clogged because we ask people to evacuate in their own low-density vehicles. While the “Cajun Navy” did incredible work, rescuing Houstonians from flooding in private boats after Harvey had passed, there is no large population of volunteer bus owners we can call upon to evacuate the general population before disaster strikes. The United States must be able to mobilize enough high-density vehicles, like busses and troop carriers, to evacuate a major city in 30 hours.

To be sure, it will be difficult to pry people from their vehicles, and leaving cars behind may very well increase the cost of property damage. Nevertheless, reducing the death toll has to be the higher priority. Communicating the importance of evacuating in high-density vehicles may be just as important a leadership task as providing the vehicles.

We should keep in mind that if the changing climate continues to contribute to these large events, we may have to be prepared to evacuate multiple regions simultaneously. Currently, in the United States alone, swathes of Texas and Florida are dealing with large-scale post-hurricane flooding, millions are without power, and wildfires are scorching lungs and clogging skies in California, Montana, Oregon, and other Western states; America’s near neighbors have also experienced a plague of natural disasters this season, with wildfires in Western Canada and devastating storm damage in the Caribbean and Mexico. To cope with future seasons like this, we will need a fleet of well-maintained, high-capacity vehicles to move evacuees who will leave their beloved cars behind. It is the only way to manage the evacuation of millions of people.

Gimme shelter. It is also imperative to provide reasonable shelter for the evacuees who will be displaced for significant lengths of time. Currently, evacuees often stay with friends and family or in hotels, frequently having to pay upfront costs. They can’t necessarily return home immediately after the life-threatening event has passed.

In 2000, I fled the Cerro Grande forest fire in New Mexico with my family, along with hundreds of others. We were displaced from our home for about six weeks during the fire and subsequent repairs. Our choices were to sleep on cots in a high school, stay in trailers provided by the Federal Emergency Management Agency, or stay with family. We were lucky: We could afford to travel and stay for an extended period with my in-laws, but that is not an option available to everyone. Certainly, as some Floridians found out when they attempted to leave ahead of Hurricane Irma, air fares can suddenly inflate when demand spikes. As a society, we should be able to provide evacuees not only with a way to get out, but also with a decent place to stay while they are displaced.

Improve damage forecasts. Weather forecasting has improved a great deal over the years, and overall, predictions regarding the path and intensity of both Hurricanes Harvey and Irma were quite accurate. However, the forecasts were not able to precisely predict which areas would suffer the most damage. In order to minimize the need for evacuation, we need more accurate damage forecasts, sooner. They have to account for all possible difficulties people may face, including power, water, and sewer outages. This will probably require a new investment by the federal government in the required technology and manpower.

Finally, we need to make sure authorities speak with one voice. In Houston last month, the state governor advised people to get out ahead of the storm, but local authorities told everyone to stay put. Of course, civic leaders have to weigh many variables before deciding whether to tell people to evacuate or shelter in place, but once the decision is made, it’s essential that they clearly communicate and implement it. Given the sheer scope of any effort to evacuate millions of people from a major costal city, and the fact that most operations of that scale will require help from multiple states, it may make sense for the federal government to take on this decision-making.

The climate is not going to stop changing soon, so I expect we will be forced to deal with more and more evacuations. The time to prepare for them is not three days before a disaster hits. Recent evacuations have not been as successful as they could have been. We have the technology to do better. When will we find the will?