Independent G.I.S. lab, Tornado Siren placement

Independent Geographic Information Systems Lab

Initial Statement..

During an advanced G.I.S. course that I took at Ball State University, we we're informed that we had to build a spatial model that should answer a specific geographic question or problem. The problem I addressed was the relation of tornado sirens to specific population densities in the county of Delaware, IN to see if adjustments are needed. As stated in the write up, I obtained a siren shape file from the Delaware county G.I.S. department and then searched the schools geodatabase to gather census data layers to depict locations of population. Once I collected the proper data, I designed the model builder to clip Delaware County out of a United States County shape file. It accomplished this task by selecting the proper area of Delaware County and then creating a separate layer. In the background, it also buffered the locations of the sirens to a one mile radius (a very crude, general estimate of how far the sound can travel especially during severe weather) and then dissolved the data to give a smooth appearance. From there I combined both siren buffer and the clipped Delaware County layers to come to depict my conclusion. Most of the town seemed to fit within the boundaries of the sirens output range except for two elementary schools. It seemed I had found a small flaw in distribution of sirens, so I discussed how a siren should be placed in close proximity of the schools to increase warning time to there locations and to protect lives.

After some critiques were provided to me by the Delaware County G.I.S. and EMS coordinators..

I soon realized that I had several errors in my thinking. Yes, my model was acceptable by Ball State standards because it did in fact do what it was supposed to do but it did not answer the deeper question I was hoping to resolve. The first problem I introduced into my project was the data for population. Census data isn't nearly as accurate as using an address point layer. The reason is, census layers use blocks (They are usually small areas bounded on all sides by visible features such as streets, roads, streams, and railroad tracks, and by invisible boundaries such as property lines, legal limits, and short imaginary extensions of streets and roads." (from: Datamocracy description)) and not personal locations like in an address point layer. So already, it was questionable about what the sirens did and did not cover. Another critical error in my assumptions was the purpose of tornado siren. It is simply to warn the general public who are in an outdoor place to move inside and to seek shelter. Not to warn people who are already indoors and near a shelter. In fact, research has been done that during severe weather and in an interior house, the sound of the siren would only be audible up to a quarter-mile at best. So overall my conclusions in the paper were wrong on a professional level but provided me with a lot of detail and insight into how warning systems are planned out. I was very fortunate that I had professionals critique my ways of thinking in a constructive manner and am thankful for it.