How GIS Can Aid Emergency Management
Susan Cutter, a geography professor at the University of South Carolina, discusses the use of GIS in emergency management and the ‘why of the where’ when working with maps.
Susan L. Cutter is a Carolina Distinguished professor of geography at the University of South Carolina where she directs the Hazards and Vulnerability Research Institute. Her primary research interests are in the area of disaster vulnerability/resilience science — what makes people and the places where they live vulnerable to extreme events and how vulnerability and resilience are measured, monitored and assessed.
Cutter is a GIS hazard mapping guru who supports emergency management functions. I posed a series of questions about mapping and asked her to respond in writing. In Cutter’s responses she reminds us to ask the “why of the where” question when looking at maps.
What has been the evolution of hazard mapping in the United States, and how does that compare with what is being done in other countries?
Hazard mapping has a long history here in the U.S. going back to the 1960s with the work of Gilbert F. White and his insistence that we map not only where the hazards are, but where people live and work relative to the risks, what he called the human occupance of hazardous areas. Hazard mapping has evolved hand-in-hand with the understanding that we can never truly control nature. Mapping has shifted from a focus on the event itself (modeling physical processes) toward a focus on understanding interactions between people and the environment. The U.S., because of the large diversity in possible hazard threats to the nation, has become a leader in hazard mapping and the integration of new tools and technologies (such as GIS, remote sensing, GPS) into the emergency management cycle.
Besides mapping hazards — such as flood zones, seismic areas and the like — how else do you see a GIS map being useful to emergency managers?
GIS is more than mapping. It is also an analytical, data management and visualization tool. GIS can be used for situational awareness, for identifying ideal locations for prepositioning assets ahead of an impact, for understanding the relationship between hazard exposure and social vulnerability as part of the hazard mitigation planning process. GIS models and simulation capabilities enable decision-makers to both exercise response and recovery plans during non-disaster times and also understand near real-time possibilities during an event. Essentially, if you have data, it can be mapped, analyzed and utilized to make better decisions in a measureable amount of time.
The best thing that emergency managers can do is identify local partners who could assist with mapping and analysis needs. This could be a local community college, college or university. One place to start that is specific to hazard assessment would be the local or state HAZUS user’s group, which would likely include qualified and interested geospatial experts focused on hazards. The other option is to work through the local planning departments and county councils of government. Rather than focusing on the master planning process, work with them on emergency management topics such as hazard mitigation and risk assessment.
There are many forms of social data — information about people such as age, income, ethnicity drawn from sources likes the U.S. census, or data derived from social media. The former is used to assess the location of vulnerable and special needs populations within a community. Knowing about the landscape of this social vulnerability helps to identify which populations may need assistance in preparing for, responding to and recovering from events.
Social data from social media is currently used to disseminate messages and information from emergency management in a top-down approach. Steps are being made in research circles to utilize “citizens as sensors” to create a more realistic real-time picture for situational awareness to aid decision-makers.
The choice of software will be a function of the resources and expertise that is locally available. If you don’t have a dedicated GIS person, it may not make sense to have the full complement of Esri software. In these instances use of “best available” data from online sources may be the appropriate choice. On the other hand, Web-mapping, mobile data collection and analysis, and desktop modeling using Esri software is now approachable and very useful for the less than hard core GIS users, like myself.
You as a university support the South Carolina Emergency Management Division (SCEMD). Do you think that is a good model for others to try to establish? If so, how would you advise states and universities to find a mechanism for partnering?
The partnership we formed with SCEMD has been beneficial — students see the real-world application of the work they are doing; SCEMD gets cutting-edge science infused into its programs and in some cases becomes a model for the nation. It is a win-win situation. Each state will have a different mechanism for partnering. Not all emergency managers may be amenable, nor will all universities. The key is identifying people who are willing to work together for a common goal and take it from there. If there is willingness to work together for the betterment of the state, mechanisms can be found to formalize collaborations.
Mobile technology like smartphones and tablets along with higher connection speeds are revolutionizing how we access data. What impact do you see this having on computer mapping, its use and how we can adapt that to emergency management purposes?
These technologies are revolutionizing emergency management and mapping as well. For example, we now collect field data on recovery using iPads and directly upload data to the cloud and our servers at the university. This technology not only cuts down on processing time (and errors) but it also means we can generate maps much more quickly. Real-time damage data could be collected using this method post-disaster, meaning that detailed preliminary damage assessments can be produced in hours, not days. Add in citizen sensor data from social media and the possibilities of crowdsourced damage and recovery information becomes a reality.
Social media continues to increase in the way it is being used to impact people's daily lives. What general uses for mapping can be applied to leverage social media use by average citizens that might benefit community resilience?
Mobile devices have geocoding within them. Currently we can, as an example, look at Twitter and see what is being said and (more importantly) where tweets are coming from on the ground. Maps of the tweets and content provide a better picture of the situational awareness, impacts and citizen status in a truly ground-up rather than top-down approach. It is an exciting and new field that is relatively unexplored at present.
You have provided some good advice as to how to use GIS to improve emergency management programs. What mistakes have you seen people make in using GIS in emergency management, and what should we do to avoid them?
One of the biggest issues has to do with cartography (the science of making maps). Just because you can use the GIS software, doesn’t mean you understand the fundamental nature of spatial (or geographic) relationships. Once they see the map, emergency managers need to ask additional questions as to why (or what I like to call, the why of the where). The map shows the distribution of shelters and occupancy, for example, but a further question is why are some shelters over-subscribed while others are not? Also GIS personnel should become aware of the pitfalls of using certain classification or symbolization schemes so as to avoid misrepresenting data or displaying data out of context.
Is there anything you would like to add?
It is important to me personally and professionally that research be used to improve the human condition by being relevant, useful to practitioners and providing the empirical basis for sound public policies. To find out how we go about doing this, y’all come for a digital visit to HVRI at www.webra.cas.sc.edu/hvri.