GIS and disasters: Planning for catastrophe
Recent events, including the Indian Ocean tsunami, the hurricanes of the 2005 season, and the 7/7 and 9/11 terrorist attacks, have made us all acutely aware of the vulnerability of modern society. Knowing where such events have occurred and the geographic limits of their impacts is clearly important, particularly when combined with information on human populations, infrastructure, and other spatially distributed phenomena that may be relevant to response and recovery. But more generally, geographic information and the technologies that acquire, interpret, and disseminate such information (GIS, remote sensing, GPS, etc.) are clearly essential in all aspects of disaster, from preparedness, prevention, and protection through detection to response and eventual recovery. Geographic information and technology (GI&T) provide the basis for estimating and mapping risk, for planning evacuation routes and shelters, for determining areas where human populations are most likely to have been impacted following a disaster, and for assigning resources during recovery, among many other vital and important tasks.
In principle, investments made over the past decade in digital libraries, geoportals , data warehouses, and the Internet that connects them and allows users to access them (see, for example, , should have created a world in which relevant data can be assembled quickly and easily following an event (Goodchild, 2003a). In reality, however, things are still far from ideal, and during the period of initial response access to GI&T is likely to be confused and ineffectual rather than smooth and efficient. The New York City emergency-management GIS was housed in one of the buildings destroyed on 9/11 and had to be re-established in the hours following the disaster. Communication links are often disabled, as they were over large areas of Louisiana and Mississippi after Hurricane Katrina. First-responders are often deluged by massive data sets provided voluntarily by agencies, corporations, and individuals in the hours following an event, but have no way of making effective use of them until power is restored, systems are up and running, and suitably trained staff are available. In reality no-one can anticipate the locations, scales, and intensities of disasters like these, or ensure that the necessary systems are hardened and that data to support response and recovery are accessible to everyone prior to or immediately after the event. Geographic data sets are often subject to licensing and access restrictions that cannot be arbitrarily ignored during disasters, and too few people on the ground are sufficiently trained to use them effectively.