At the end of 2012, I deployed with Disaster Tech Lab to the New York City area to respond in the aftermath of Hurricane Sandy, the second-costliest hurricane in U.S. history. And earlier this year, super-storm Darwin hit Ireland, where I live, causing widespread damage and flooding.
Disaster Tech Lab has responded to many other disasters since I founded it four years ago after responding to the 2010 earthquake in Haiti, but my experiences in New York and here at home after Darwin prompted me to think about how information is collected and distributed in disaster zones.
Following the initial rush to extol the benefits of “big data”, many have realized that although data gathering is important, the curation and dissemination of relevant data is really where the rubber meets the road, especially in the aftermath of a disaster.
Well-organized data distribution allows not only for a better tailored response effort but also increases awareness among the general public of areas that may be dangerous or lacking in services, and gives the public a central access point for data on response efforts as well as where aid is available.
Following super-storm Darwin, I conducted an on-line survey on how information on storm and flood damage was being disseminated to the Irish public. The survey was basic and aimed only to ascertain if people had received information about damage caused by the storm and through what medium they had received the information.
Seventy-eight percent of respondents reported they had encountered flood damage during travels around the country following the storm, but only 56 percent had been aware of the damage prior to seeing it. Most interesting to me was that 83 percent received this information via social media while only six percent received the information through local authorities.
The Irish local and national authorities appeared to have been woefully unprepared for an event that was obviously not without precedent. The only publicly available resource was the floodmaps.ie website run by the Office of Public Works, on which the most recent data dated from 2011 and was gathered through a retrospective study. There was no obvious way to report damage in real-time.
This was a shame, because there are a number of easily implementable, open source platforms that allow the creation of an online map with a real-time reporting mechanism without having to re-invent the wheel. Such a map allows people to report any flooding or storm damage they encounter via email, SMS or the website itself. Reports can include the reporter’s location (automatically mapped), description and still images or video.
The data can serve dual purposes. It can alert response services to issues that require intervention or remediation, and it can inform the general public about flooded roads, downed trees and other issues that affect transport and other public services. With this information, citizens can plan their travel to avoid problem areas, with the result, we hope, that traffic flows more fluently.
Once curated (again in real time) this information can then be displayed on an online map. In my view, curation is a critical step in developing a reliable, useful information resource. Unfortunately a lot of disaster data that is aggregated and distributed via social media nowadays is accidentally or purposely incorrect.
And as you know – you do know this, don’t you? – “I saw it on the Internet” is not necessarily proof that something is true. Incorrect information that is shared virally can cause significant problems in a disaster zone, but some organizations go too far the other way, trying to “control the message” by telling people how to tweet or post on Facebook during disasters. These sort of efforts are counterproductive as they can result in the production of filtered data that ignores important problems and potential solutions.
In the absence of a government resource after super-storm Darwin, Disaster Tech Lab created the Irish Flood Alert Crowdmap, which allows people to report the location of floods and flood damage and to upload images.
Following Hurricane Sandy, Disaster Tech Lab created a crowd map that allowed citizens to report and learn about communications network outages and other service issues. Our Sandy response was our first in a developed world setting, and although pre-disaster communications networks were much more ubiquitous than in previous theaters where we had deployed, and the responses from telecommunications providers were robust, responders still faced challenges in knowing where communications services were down, and more important, where they were still working or had been restored.
Interestingly and somewhat shockingly, these challenges were in large part the result of telcos not sharing data. Not only among themselves, but also with the Federal Emergency Management Agency (FEMA). As a result, crowdsourced data became the most reliable source of information for the general public.
The same was true in Japan in 2011 following the meltdown of nuclear reactors in Fukushima, after an earthquake and tsunami devastated the country’s northern region. In a serious absence of reliable radiation data, and with mistrust of government extremely high, a volunteer organization called Safecast created incredibly detailed maps of radiation levels all over the country.
It’s clear we will see (are seeing) a drastic increase in extreme weather such as floods and storms, and a shift from disaster response to disaster preparedness is needed. Increased public awareness – empowering people by giving them a means to report problems and needs in real time and act on this information – will lead to increased preparedness and resilience.
Evert Bopp is a Dutchman who has made Ireland his home. After 15 years in commerce- and revenue-driven fields in the tech industry, he founded Disaster Tech Lab in 2010 to provide rapid response communication networks for use in disaster relief and humanitarian aid work globally by responding organizations, as well as affected communities.