Large-​​scale emer­gen­cies, such as bomb­ings and plane crashes, trigger a sharp spike in the number of phone calls and text mes­sages sent by eye­wit­nesses in the vicinity of the dis­aster, according to a research study by net­work sci­en­tists at North­eastern University.

The find­ings, reported in the online-​​only sci­en­tific journal PLoS ONE, could trans­form the ways in which real-​​time com­mu­ni­ca­tions tools, such as smart phones, help policy makers and emer­gency per­sonnel respond to poten­tial tragedies. The social net­working web site ushahidi​.com, for example, became a very pop­ular tool for tracing the needs of vic­tims of last year’s earth­quake in Haiti.

Our work may have impor­tant impli­ca­tions for policy-​​makers who want to rethink how emer­gency response tools are used,” said coau­thor James Bagrow, a post­doc­toral research asso­ciate for Northeastern’s Center for Com­plex Net­work Research. “The sheer objec­tivity and volume of our data could help save lives.”

Albert-​​László Barabási, director of the Center for Com­plex Net­work Research, and Dashun Wang, a PhD can­di­date at the center, also con­tributed to the report.

The researchers ana­lyzed anony­mous billing records of 10 mil­lion mobile phone sub­scribers in a western Euro­pean country from 2007 to 2009. They com­pared call activity in the imme­diate after­math of eight unplanned emer­gen­cies with eight sched­uled activ­i­ties, including rock con­certs and sporting events.

Bomb­ings and plane crashes—the most threat­ening disasters—elicited the greatest spike in call activity, as well as the most rapid decline in call volume. Con­certs and sporting events, on the other hand, induced a more gradual increase and steady decline in call volume.

People demon­strated an urge to use a cell phone as a response tool imme­di­ately only after extreme emer­gen­cies,” said Wang, who noted that eye­wit­nesses tended to call mem­bers of their social net­work within min­utes of the anomaly.

News of the most dan­gerous events often spread quickly and effi­ciently from an eye­wit­ness to indi­vid­uals as many as four links removed from his imme­diate social con­tacts, said Bagrow. Less threat­ening emer­gen­cies, such as minor earth­quakes and black­outs, showed little prop­a­ga­tion beyond the imme­diate social links of an eyewitness.

Infor­ma­tion spreading is actu­ally very rare,” he said. “This means that a population’s innate ret­i­cence to com­mu­ni­cate may nat­u­rally sup­press false infor­ma­tion and may explain why the dis­aster myth” — the belief that panic is a common, wide­spread reac­tion to an emer­gency — “con­tinues to hold, even with today’s con­stant communication.”