Lesson 3: Analyzing Spatial Data in a GIS

The purpose of this exercise was to follow an emergency relief effort in response to tornadoes that struck Oklahoma on May 3, 1999.

The green stars indicate the priority relief sites (schools, churches, and hospitals) within a specified distance of the tornadoes, and the darkest colors indicate the most important areas in terms of priority relief.  Listed below are the relief sites as selected in my attribute table:

There are obvious limitations to this analysis as far as 'real-world' emergency relief planning goes.  One of the biggest assumptions is that population density is going to be uniform throughout each county, and in a state like Oklahoma, that is simply not true.  Population density data would have to be calculated individually for each census tract, rather than across the county, since many counties in Oklahoma typically have one county seat where most of the population is located.  Thus, priority could be valued incorrectly by classifying areas with a "higher" population density as a higher-priority relief zone, if there's only one big city in a county with not many rural cities.  Adding a roads data layer would definitely help to augment the other data layers, because it would add another element of determining priority zones by showing potential evacuation and transportation routes.  It would also make it a bit easier to show routes for sending relief supplies.  Overall, there are special circumstances to every type of natural disaster, which makes this a difficult type of study to accept as a complete and final strategy for emergency relief.