What is the best way to re-project NAD83(86) to NAD83(2011) without changing the x y coordinates? We have a been processing survey data for the last few years from text files that are referenced to HARN, CORS96, 2007, or 2011 (presumably what local surveyors have used over the last 10 years with network GPS receivers) and importing them directly into our NAD83(86) enterprise geodatabase. We would like to re-project our enterprise geodatabase to NAD83(2011), but maintain the x and y location as these represent the actual survey data.
Initial testing has shown that re-projecting from NAD83(86) to NAD83(2011) in ArcMap does not change the coordinates considerably (eg, <0.0001'). This is confusing as I would have expected about a 0.9' shift based on limited analysis on local survey monuments, the biggest delta between NAD83(86) and HARN.
What is the best practice approach to this problem?
If you don't want to change the coordinate values, use the Define Projection tool, or the data's property page in ArcCatalog.
The offset will depend on which transformations were used. Did you set any? You also might have to do an actual conversion in steps like NAD 1983 to NAD 1983 HARN, HARN to NSRS2007, NSRS2007 to 2011. The latter ones may need 10.4 and the ArcGIS Coordinate Systems Data setup.
What area(s) does this data cover? And I can suggest some transformations.
I appreciate your expertise on this subject and frequent contribution to this forum. Our organization has for some time been collecting and storing data in SDE as "NAD_1983_StatePlane_North_Carolina_FIPS_3200_Feet" (WKID 2264). Is that the same as NAD83(86)? We are about to receive updated high resolution aerial imagery for our region (coastal NC) that is supposed to be delivered in NAD83 2011. Am I right in thinking that it would be optimal to convert all of our legacy data to match the updated basemap imagery? If so, what process do you recommend for doing so?
City of Wilmington NC
Thank you for the kind words.
This gets interesting. Technically, the "NAD 1983" geographic coordinate system should be the original, 1986 version. However, a LOT of people have used it for every other possible re-adjustment. Unless there's metadata/history, it can be very difficult to determine which data is really HARN (and there might be multiple versions in some states), NSRS2007, etc. The conversions between NAD83 (86) and HARN will be the largest, with the HARN/NSRS2007/2011 conversions much smaller, cm-level usually.
You might want to run some test points in various areas through each transformation, NAD_1983_To_HARN_NC, NAD_1983_HARN_To_NSRS2007_GEOCON_CONUS, and NAD_1983_NSRS2007_To_2011_GEOCON11_CONUS.
Depending on the data accuracy, the transformations, particularly the last two, may not really matter. As I mentioned earlier, the GEOCON transformations were added at 10.4.1, and need the separate ArcGIS Coordinate System Data setup.
That is helpful, thanks Melita. What we have typically done over the years is to use various generations of aerial imagery, beginning in 1994 and every 4 years after as a base for collection and update of planimetric data. Until the last 10 years, we haven't had very good metadata on the imagery, but we have just called everything "NAD 83" whether that is correct or not. So, based on that, we should just convert what we have from (presumed) NAD83(86) to HARN and call it good?
If you don't mind, can you elaborate on the process of running test points in various areas through each transformation? I'm not really clear what that means or how it's done.
Thanks very much for your input.
Yes, particularly with imagery which has its own resolution, that would likely work (convert to HARN).
Identify some coordinates at the edges of your area of interest, plus some interior points. Try to include points at the low and high points as well. I'm pretty low-tech. Stick them in a text file like:
Add the file to ArcMap with the Add XY Data option. Make sure to set the coordinate system to NAD 1983. Right-click the layer to export out a shapefile which we'll call nad83input.shp. Use that in the Project Tool, set the output coordinate system (CS) to NAD 1983 HARN. Use the NAD_1983_To_HARN_North_Carolina transformation. Set the output shapefile's name to nad83harn.shp, and so on. If you add them all to ArcMap but do not set any transformations, you'll be able to see the offsets. You could also add the coordinates into the attribute table and then copy them to Excel to automatically calculate the differences. Or keep adding fields to each new shapefile and calculate the current coordinate values. Then you can see how the data's trending between the different re-adjustments.
Particularly with the HARN to NSRS2007 and the NSRS2007 to 2011 transformations, NGS wouldn't create transformation grid files because the standard deviations were about the same size as the offsets themselves. They did find that it wasn't just noise so a few years ago, they did create grid files. There are newer ones in GEOCON v2, but we haven't implemented them yet.