If a zombie apocalypse breaks out in the United States, where should the “zombie-fighting headquarters” be located? Oh sure, a zombie survival strategy may not seem important now, but it’s official research that will be much appreciated when the nation is infested with hungry-for-brains undead. Cornell University researchers asked, “If it takes a month to develop and distribute an effective vaccine (or an effective strategy for zombie decapitation), what regions should one locate the zombie-fighting headquarters?”
Cornell University researchers presented “You Can Run, You Can Hide: The Epidemiology and Statistical Mechanics of Zombies” (pdf) at an American Physical Society meeting in San Antonio last Thursday. “We ran full scale simulations of a zombie epidemic, incorporating each human in the continental United States, and discussed the geographical implications for survival.”
“We all know that a zombie infestation is inevitable, so why not try to understand the effects on the US with a simulation?” Matt Bierbaum, a Cornell PhD candidate in physics, said, “We made such a simulation for the web in which all 308 million people from the 2010 census interact locally in the US. The dynamics are Gillespie dynamics in which every possible interaction is considered at each time step.”
So how long would you have to survive after the nation suffers an infestation of the undead? You have about one month, or 28 days. Hmm…haven’t we seen that movie?
“After four weeks, much of the United States has fallen,” the researchers wrote, “but it takes a very long time for the zombies to diffuse and capture the remaining portions of the United States. Even four months in, remote areas of Montana and Nevada remain zombie free.” By that reasoning, it would seem like hiding in the most remote areas of Alaska might keep you safest from the zombie infestation risk.
If you live in New York City, you should make a run for the northern Rockies, advised Alex Alemi, a PhD student at Cornell University's Physics Department. The movies got it wrong. He said, in a film or book, "if there is a zombie outbreak, it is usually assumed to affect all areas at the same time, and some months after the outbreak you're left with small pockets of survivors. But in our attempt to model zombies somewhat realistically, it doesn't seem like this is how it would actually go down.”
Inspired by zombie movies, the researchers chose parameters that are reflected in such films. They pointed to previous research that was modeled after Night of the Living Dead and Shawn of the Dead. The bite-to-kill value was 0.8, meaning “zombies in the films are 1.25 times more effective at biting humans than the humans are at killing the zombies.” They then adopted that value for their simulation. The research paper states, “Finally, to set a rate for our bite parameter, we similarly assume that the zombies are undergoing random motion inside the cell at 1 ft/sec, and they interact with a human anytime they come within 100 feet.”
The researchers’ simulation depends upon the SIR epidemic model, which is used to predict the spread of disease in real life. But the SIR model has different variables than could be applied to zombies, because people are either susceptible, infected or recovered. Yet recovery isn’t going to happen because “once a zombie, always a zombie.” Alemi said, “Zombies don't get better, nor do they die, so the only way you can get rid of a zombie is for a human to actively kill it.”
So their simulation model is based on four states anyone could be in during an outbreak; a person could be a healthy human, an infected human, a dead but roaming-and-hungry zombie, or a dead zombie. They ran many simulations, including “7,000 different 28-day zombie outbreaks in the continental United States starting with a single individual.” They even studied an outbreak in a tiny town "with an initial population of 200 people, 199 uninfected and 1 infected.”
Alemi added, "Each possible interaction--zombie bites human, human kills zombie, zombie moves, etc.--is treated like a radioactive decay, with a half-life that depends on some parameters, and we tried to simulate the times it would take for all of these different interactions to fire, where complications arise because when one thing happens it can affect the rates at which all of the other things happen.”
Surviving a zombie epidemic would require more than simply avoiding densely populated urban areas. The researchers discovered that the probability of surviving a zombie outbreak is not just linked to population density, but also to your distance from major cities. They wrote, “After 28 days, it is not the largest metropolitan areas that suffer the greatest risk, but the regions located between large metropolitan areas.” They found that “the area with the greatest one month zombie risk is north eastern Pennsylvania” where population density is not so much of a problem as its location to major population centers.
"Given the dynamics of the disease, once the zombies invade more sparsely populated areas, the whole outbreak slows down -- there are fewer humans to bite, so you start creating zombies at a slower rate," Alemi explained.