Map projections and methods for measuring road distances
Our transportation modelers want to make sure that the road network shapefile they use has the most accurate distances possible. They are concerned about potential distortion in our usual coordinate system (GCS_North_American_1983) and the inability to use the "Calculate Geometry" function on a geographic coordinate system. Right now the typical procedure is to set the data frame properties to a projected coordinate system (NAD 1983 UTM Zone 15N) and then run the distance calculation using the projected coordinate system. We have no idea how much error we are introducing. Does the "Calcutate Geometry" function compensate for the varying amounts of distortion across an extent? Or is the difference in distortion negligible at the extent we are working with? Is there a better method to get accurate road segment lengths? Should we be using a different projection, such as an equidistant one?
Our region comprised of eight counties roughly centered on St. Louis, Missouri, with an extent bounded by: Top 39.000000 dd; Right -89.595230 dd; Bottom 38.003499 dd; Left -91.369228 dd.
Mapping Center Answer:
Of course the amount of error you are introducing depends on a number of factors, including 1) how much area you are mapping, 2) the location of you are mapping relative to the map projection's parameters (e.g., central meridian, standard parallels), and 3) how much error is "too much".
In general, large scale maps are those that can be used to make accurate measurements; these would be at a scale of 1:250,000 or larger (4" or fewer to the mile). On a large scale map, the change in scale across the surface of the map is negligible, so you can trust the map to be an accurate representation of the small piece of earth that it covers.
A UTM zone is designed to be no worse than 1 part in 2500, and we note that you are crossing outside the zone (-90 is the right edge of zone 15).
There is also some error creeping in because the spheroid (ellipsoid) surface of the datum is not at average elevation. Some places have published county-level coordinate system that take average elevation into account.
Keep in mind that other factors can also affect the map accuracy, including error in the data themselves, error in the tracing method used to describe a route, and even user error. Chapter 2 in Map Use: Reading and Analysis, Sixth Edition describes issues related to map scale and Chapter 10 discusses Map Accuracy. You might find reading these chapters helpful.
If you would like to post a comment, please login.