During the Illinois floods of January and February 2008, I volunteered again with the American Red Cross (ARC), and found myself working in Disaster Assessment (DA), the ARC function that gathers information to help ARC respond to the needs of the affected population. Relief operation managers need to know the extent of the disaster, the number of people, and the degree to which they are affected, but also the weather, road conditions, power outages, business closures,… anything that influences the state of the community and the response to their needs. Since I’m a hydrologist, I knew of lots of Internet data useful in responding to a flood-caused disaster: FEMA Flood Insurance Rate Maps for the extent of 100-year floodplain, TerraFly for demographics, Google maps to check addresses, and digital elevation models to estimate the inundated area using NOAA-USGS river stage. In the end, we were able to get enough information out to help keep the disaster relief operation running, and got high marks for our contribution.
But when we were done, I knew there was room for improvement. The people conducting DA used Google maps, but virtually nobody else did. ARC operation headquarters had DeLorme mapping software but DA staff in the field did not, so neither side could send a map that could be zoomed or edited. We made do with faxes and couriers, leading to some unnecessary expense as well as minor confusion over the extent of the disaster and progress on surveying the damage.
It is a truism in disaster management that successful response starts with a map, so it isn’t surprising that some of earliest adopters of geographical information systems (GIS) were emergency management agencies. GIS software combines map illustration with databases of spatial attributes, allowing the user to query and plot spatial data (e.g., show shelters that have generators and lie outside the 100 yr floodplain). Although there are freeware GIS, the most widely used software is ESRI’s ArcGIS, which can be found in many State EOC’s (and in ARC national headquarters). Unfortunately, ArcGIS is expensive, complicated, and uses a propriety data format, leading to data stuck in silos and understaffed EMA mapping teams overwhelmed during a disaster. In my disaster experience, ArcGIS maps arrive 5 days after they would be useful for ARC, if they arrive at all. On this particular disaster relief operation, I didn’t see any ArcGIS-generated maps.
One mapping solution for ARC DA is Google maps, a web-based service for creating and sharing maps, with a free license to its Application Programming Interface (API; this is geek-speak for “it’s programmable”). Google maps are being incorporated into ARC training manuals for DA and are embedded in RDMS, the GRT surveying/database system adopted by ARC for Detailed Damage Assessments:
Unlike ArcGIS, Google Maps requires no special training or license, and is available anywhere there is an Internet connection. Unlike Google Earth, there is no software to download or install, one simply signs up for a My Maps account and begins creating, saving, and sharing customized maps. Google My Maps is particularly powerful for DA because it is easy to use and shared maps can be edited by multiple collaborators simultaneously to create a real-time situation map. Google My Maps chief advantage over ArcGIS is that it brings the information out of the EOC silo to people in the field, allowing them to add information on current conditions. A good example of Google maps used for a disaster response is the map of shelters and fire locations for the 2009 Jesusita Wildfire near Santa Barbara, CA.
Even with all the features of Google maps, it was still hard to see how the average ARC volunteer would repeat my trick of querying census data from Terrafly to estimate the effected population. To make census data more accessible, I put together the Disaster Assessor, a mashup of Google maps, census data, and floodplain maps to assist in estimating the population effected by a disaster. The Disaster Assessor allows the user to draw a polygon representing the disaster footprint and estimate the effected population; the polygon and the result can then be imported into a Google My Map for anyone to see or edit. The Disaster Assessor is built using public domain software and data (Google maps API; US Census 2000; and FEMA floodplain insurance rate maps). Putting it together required about the same effort as any web site, plus a little coding and reformatting of the data. While there are certainly opportunities to enhance the interface, extend coverage to other counties, and add functions, the Disaster Assessor shows how one could build a lightweight GIS for disaster assessment. And it’s free.