Madame Bertil, 88, packed up what little she had left in the one-story house her son had built for her and prepared to leave Port-au-Prince. It was 21 January 2010, 8 days after a magnitude-7.0 earthquake ravaged the capital of Haiti, killing between 65,000 and 300,000 people and leaving over 1.8 million people homeless. Scared and unsure, Bertil called her brother Luca and made arrangements to stay with him almost 160 km away in the district of Acul du Nord. It would be 7 months before she returned to the capital. Bertil was just one of the more than 630,000 Port-au-Prince residents who fled the city—some leaving with nothing, not even shoes—in the aftermath of the earthquake.*
Even though so many people had left Port-au-Prince, relief efforts were focused only on those who stayed; those who left had to find help on their own. That may be less of a problem in the future, thanks to a new analysis of cell phone records from the Haiti quake. The data has enabled researchers to study how populations move after disasters, a finding that might help aid agencies better target their efforts.
Large population movements, such as those after the tragedy in Haiti, are not uncommon. Every year, tens of millions of people abandon their homes because of natural disasters. People who leave disaster zones often do not receive the aid they need because relief agencies either don't know where they are or that they even exist. Such was the case in Haiti where the already poor, rural towns outside of the capital had to provide care to the fleeing masses—almost 23% of Port-au-Prince's total population—without the necessary outside help.
Epidemiologists Linus Bengtsson and Xin Lu from the Karolinska Institute in Sweden wanted to figure out where people went after the quake. The researchers examined the movements of over 2 million anonymous mobile phone users by collaborating with Digicel, the largest mobile phone operator in Haiti. Digicel provided them with the locations of unspecified individuals for a time period spanning from 42 days before the earthquake to 341 days after it. Calls made on people's phones represented their locations. Every mobile phone has a SIM card that registers so-called chargeable events—an action that the phone company can bill them for such as a phone call. Chargeable events connect with the mobile phone network through nearby mobile phone towers. Once the network database registers the call, it records which tower made the connection. Towers can vary in their coverage areas from 5 square kilometers to 100 square kilometers. By following these connections, the researchers tracked people's locations as they left Port-au-Prince.
Despite the chaos, the team reports online today in the Proceedings of the National Academy of Sciences, the movement of displaced people was so regular that the researchers could predict with 85% accuracy an average a person's location during the first 3 months after the earthquake .
The team found that after the earthquake, "people seemed to have traveled to where they had their significant social bonds and support," says Bengtsson. Specifically, Haitians went to the same locations where they had spent Christmas and New Year's. For example, departments (Haiti's major administrative divisions) Sud and Ouest received the highest influx of people. By understanding where people go during their holidays, the researchers say that they can accurately guess where people will go during times of disaster.
According to the researchers, this information is relevant for future relief efforts for two reasons. "First, it gives important information about where do people go after a disaster and when do they come back, and second, it shows that people's locations during a disaster are highly related to where they traveled normally," Bengtsson says. The team proposes that by understanding these patterns, relief agencies can deliver aid more rapidly in the future and to the right places.
Their findings also showed that people slowly began returning to Port-au-Prince beginning in February. The return rate picked up as time approached into April, after which it leveled off and people came back at a stable rate. But the findings do not explicitly say why they returned, notes Mark Schuller, an anthropologist at York College of the City University of New York who has been doing research on globalization, nongovernmental organizations, and disasters in Haiti since 2002, but who was not involved in the current study. He says that many Port-au-Prince residents originally came to the city from rural towns in the mid-1990s in search of economic and educational opportunity. But after the devastation, many went to live with their rural cousins in search of clean water, sanitation, and help. There was little if any aid delivered to the surrounding towns, however. Many of the people returned to Port-au-Prince because they felt as if there was no life for them in the impoverished rural regions of Haiti, and they no longer wanted to be a burden on their already poor family members, he says.
Vincent Blondel, an applied mathematician and networks researcher at the Université Catholique de Louvain in Belgium, says that although the mathematics in the work are not new, applying the data set to help increase relief efforts is a new idea. He adds that taking this approach toward data sets in disaster-prone places such as Africa and India—where technology, infrastructure, and access to water and health services is limited but many people have access to cell phones—could help create better contingency plans.
Although the idea is to use this research to enhance future relief efforts, it has some limitations, notes Richard Garfield, an epidemiologist at Columbia University. This information is useful when applied to small, low-income countries, he says, but it would not have been effective in figuring out where people went after a catastrophe such as Hurricane Katrina. Victims of Katrina fled to a lot of different places, including Georgia and Houston, even as far as Maine and Seattle. A lot of their movement was dictated by government relief efforts such as buses that drove them to specific shelters, so people ended up in many places that they had never been to before rather than with their families located outside of New Orleans.
On the other hand, Garfield notes, when boiled down and applied to the appropriate situations, the new study "confirmed a piece of common sense that we haven't been able to confirm in the past: that's when things are tough people go to where they are used to going."
*This story was told to the reporter by documentary film maker Francks F. Deceus, who is filming the recovery of Haitian citizens in the years following the earthquake.