bhoen@lbl.gov

Estimating Size of Error using one US projection to calculate distance between points

Discussion created by bhoen@lbl.gov on Sep 24, 2012
Hi all,

I have two sets of points (one - set A - is ~ 200,000 points and the other - set B - is ~ 20,000) spread across the continental US.  I would like to calculate distance from all the points in Set A to the nearest point in Set B and put in a few of the attributes.  A spatial join works well for this.  I have them both projected in North_America_Equidistant_Conic.

The distance (in meters) that is calculated has some error, that might be less if I was to break the data up into smaller groups and re-project.  This is not efficient, but choosing NOT TO might introduce too much error.  I can tolerate say, 30-50 meters of inaccuracy.  Do you think my error will be less than that?  Is there a way to substantiate/calculate/estimate this error?  Any academic papers that I could reference on the subject?

Thanks, in advance, for any help.

I should add I am a novice and would not feel comfortable using R, python, or doing anything other than simple programming.  I am running ArcMap 10.0 in Windows 7 (64 bit).

Ben

Outcomes