First you identify exposure geographically. Then you examine the distribution and characteristics of the population and built/unbuilt environment within the exposed area. What then?
Andrew Cuomo, the Governor of the US state of New York, is making a climate adaptation plan for people and structures in high exposure areas of New York City. He wants to buy homes and properties, raze structures and prevent all future development on these lands, leaving them fallow or using them as buffers for the impacts of storms, flooding and other climate-related hazards.
This kind of relocation is being discussed around the world. Doing it depends on many things -- one is having good information on the distribution of vulnerability within high exposure areas.
One of the major themes of the manual is that using overly large spatial units -- or not disaggregating at all -- can lead to mis-interpretation of data. (See Chapter III for an explication and example out of Viet Nam.) Today we see an interesting, if substantively entirely different, example from professional basketball.
When NBA players shoot the basketball, they make the shot on average about 44.7% of the time. However, it turns out that where they take the shot from matters critically -- so critically that the only place they actually exceed that average is within 7.5 feet of the basket. Everywhere else is below that non-spatially-disaggregated average.
So is it wrong to quote the 44.7% figure? No. But, depending on how you use it, it could certainly be misleading.
The same is very much the case for areas with high spatial inequality, whether cities or countries. The overall figure may be masking a great divide that emerges spatially, one that is essential to understand and integrate into planning.
We are more than a decade into the GIS explosion. But what this explosion has shown is that, as intuitive and appealing as visualization of data on maps is, and as easy as it is to layer data on maps, what is hard is analysis.
There are some excellent tools on the web that layer climate, economic, social and population data together. To take just two:
These put together a wide range of important data in very useful ways. However, particularly at local scale, it remains difficult to generate an integrated analysis of these data that produces rigorous and useful results without a major investment of time and energy from a skilled analyst. Certainly, some existing tools do so, albeit in constrained ways (Google traffic, for instance). But outside of the private sector, and in ways that are accessible around the world and relevant for development challenges, not so much.
I think this is the next key challenge for using spatial data.