I used the Analysis -> Proximity -> Near tool to calculate the nearest distance between:
- a Point feature class as Input
- a Line feature class as Near
Whatever the GCS of the Input feature class (Geographic NAD27, Projected NAD27 Albers Equal Area Conic), or the GCS of the .mxd project, the reslt is always the same. And it looks as being Meters if I double-check with the Measure tool.
The only diffrence I can output is if I use Geodesic or Planar option, which shows a little difference... (and looks logic).
My question is what is the unit of the calculated nearest distance? Looks like to be Meters, but why is it that by default?
And how is it possible to set another unit for nearest distance?
A GCS is a geographic coordinate system aka the units are in decimal degrees. A PCS is a projected data set and for most of the world, the projection used will transform space so that position (and hence distance) are recorded in meters. undoubtedly you have a projected coordinate system. The method of calculating distance using geodesics will result in measures being made in meters having performed the calculation using decimal degree locations, determining the 'great circle' distance (if you like), then converting the distance measure to human readable form ... meters. A planar distance determined from projected coordinates using Euclidean distance calculations. The results of geodesic vs planar will be very similar for locations close together. The differences 'may' get larger depending on the projection used, where you are on Earth and how far apart the locations are.
In short, if you are working 'locally' versus 'globally'... use projected data. If you really need the distance along the geodesic, then use it...
The units will be same as the linear units of the input dataset. If the units of input is in meters, the output units will be in meters. If the units of input is in feet, the output units will be in feet.
You can calculate the distance in desired units by multiplying the distance field with appropriate factor (using Field Calculator).
Thanks for these answers.
The Input feature class, being either Projected (NAD27 Albers Equal Area Conic) or Unprojected (Geographic) NAD27, gives both the same results in meters.
This default unit is determined by the linear units of the Input feature class in this example. If the Input feature class is not projected (just in NAD83), I believe the default is meter, probably linked to the regional setting of the ArcGIS installation and/or Windows installation/machine.
I made another test with the Input feature class in Projected (NAD27 / StatePlane_Texas_North_FIPS_4201) and the results were in US feet, just as the linear units.
All clear now.
Having similar problems. All the features involved are in the same projected coordinate system, having linear unit feet, and they are all shapefiles. Data frame coordinate system is also the same as the shapefiles. Distances are off by varying factors ranging from 50 to 5000. I gave up some time ago on using this tool, unless there's a very large number of distances involved, due to inexplicable results like this, but maybe I'm missing something?
It looks like it is based on some flavor of mercator projection, Here's the coordinate system info form the properties window:
Projected Coordinate System: NAD_1983_StatePlane_Florida_East_FIPS_0901_Feet
Linear Unit: Foot_US
Geographic Coordinate System: GCS_North_American_1983
Prime Meridian: Greenwich
Angular Unit: Degree
What do you mean by "Distances are off by varying factors ranging from 50 to 5000"?
In your example above, I would believe nearest distance with the Near tool would be US feet... If you were using the Measure tool set to US feet and mode Geodesic and/or Planar in this case because shapefile is projected, you should get similar measurements. Is it the case ?
If you get 5000 US foot difference, it looks like a lot!