0

Possible Duplicate:
What is the most accurate coordinate system for calculating areas of polygons?

I'm currently involved in comparing the "official metropolitan" areas of several cities against their actual overall urban footprint. The official metro areas' areas can be found in other sources, but I need to get the most recent urban footprint area.

I've found the fastest way to do this is to draw a polygon in Google Earth around a city's urban sprawl. As I want to create an explanatory graphic, I then brought my GE polygon into ArcGIS 10, where I converted it from KMZ to layer, and then saved the layer data as a new feature class in my specific city/map geodatabase.

Now, out of habit, b/c I'm focusing on a specific small area, I'm projecting things in UTM. I checked the attribute of my new "urban footprint" feature class polygon and found there to be an ~600,000m2 difference between the polygon area in Arc and the same polygon's area given in Google Earth.

In light of this, my question is which area value is more accurate and should I use in future calculations I'm going to do?

3
  • 1
    When computing areas, use--of course!--an equal-area projection. For more about this, see gis.stackexchange.com/questions/20054/…. In fact, it's really the same question, so to keep this site organized, I will vote to close the current one (which will leave the two linked).
    – whuber
    Commented Jan 15, 2013 at 19:37
  • However, if I'm only calculating the area for a small region (i.e. a single city), it would also be acceptable to project the polygon in that specific UTM zone, no? If I remember correctly, the point of UTM is to minimize all 3 factors to a usable extent within a much reduced region/zone.
    – Wade
    Commented Jan 15, 2013 at 20:41
  • 1
    Yes, assuming you don't mind the built-in 0.04% range of error (which in most applications is not a problem). But if you're projecting the polygon specifically to compute its area, and not for any other purpose, why choose UTM? You might as well use an appropriate equal-area projection.
    – whuber
    Commented Jan 16, 2013 at 0:30

1 Answer 1

3

Unless the measurement you're making in GE is adjusted, it will give you an incorrect value. Take a look at this Esri blog posting about measuring length and area in Web Mercator. The scale distortion for Mercator based on a sphere (which web Mercator is) varies as 1/cos(latitude) (thanks to Melita Kennedy)

1
  • +1. And note that the area distortion is the square of the scale distortion.
    – whuber
    Commented Jan 15, 2013 at 19:36

Not the answer you're looking for? Browse other questions tagged or ask your own question.