Let us suppose that we need to calculate the "hillyness" of an area near a subway station. Perhaps the streets are covered in ice, as this sometimes happens in a New York City winter, and we would like to give priority to those hilly streets near mass transit for salting. In this experiment, I located subway stop locations using a subway stop file and I made a map using ARCGIS. I prepared a "buffer" around each station of let us say, .25 Mile. I did this because if I used a point feature, I could get only one elevation reading. What I wanted to show is the variation of elevations in a .25 mile radius around a station. If the terrain was flat, my contour numbers, which indicate elevation in feet would not vary that much. If the terrain was very hilly, such at the Dyre Avenue Station, my contour reading would vary from 4 feet to 112 feet. By calculating the Standard Deviation, I could theoretically get a value of how various the elevation readings were in my .25 Mile buffer. I produced a report in ARCGIS in which I show for a particular station the low elevation reading, the highest and the average reading and the standard deviation. I produced a report for a few of these NYC stations and please let me know if my technique is incorrect. For the Dyre Avenue example, I used 176 contour readings. This is just an experiment and I do not vouch for these numbers. In doing the "spatial join", I told the computer to use contours that intersect my buffer circle of a .25 mile. This would not count contours that are "nested" in the .25 mile radius completely and would tend to artifically make the standard deviation lower. This is only a test of the technique. Note, smaller the buffer, the higher the chances that nested contours would be counted. Any suggestions out there?
No comments:
Post a Comment