3

I have a shapefile of survey control points (WGS 84) from the public domain provided by my local municipality. I'm using a GeoXH with a zephyr antenna. I have my coordinate system set to WGS 1984. I'm getting 4.5" accuracy (post processed) when I'm recording my test point.

I then take this data back to Trimble Pathfinder Office, post process it using a station about 3 miles from where the points were taken. Pathfinder Office is configured to have WGS 1984 coordinates. I then export this data to a shapefile (without a projection).

Using QGIS (2.18) I start a new project, select WGS 84 (EPSG: 4326), and load my city provided data. I then load my test points (3 taken on tripod over one of the city's monuments). I then load my shapefile, selecting WGS 84. And my data is off by 5 ft (west, and a little north).

Any idea what I'm doing wrong?

  • Is your data5 ft off, or maybe the city data? – AndreJ Jul 22 '17 at 9:35
  • Since these are survey control points, and I'm in a fairly large city, I'm assuming that as a more or less beginner I'm doing something wrong. – Ben Jul 22 '17 at 11:16
  • 1
    Try adding a baselayer to your map. This will help you figure out if the test points are off, or if the survey control points are. – csk Jul 22 '17 at 18:00
  • @csk - What I've found is that every baselayer (OpenMap, Google Satellite, Google Street, Bing Satellite, and Bing Roads all align a little differently. I'm using the OpenLayers plugin. In general, they city's survey points, streets sidewalks, and buildings layers (all of which align perfectly to each other) are closer to the public imagery than my points. Over the next few days, I'll recollect my data to confirm that this consistently off. – Ben Jul 23 '17 at 12:29
  • Recollecting is a good idea. Depending on the visible satellites, your coordinates may be off. You can not trust the calculated accuracy. – AndreJ Jul 23 '17 at 18:43
3

One thing to watch out for when doing any RTK or post-processing is what coordinate reference system is being used by the control network? The control network may return ITRF2008 (or another ITRFxx system), which is equivalent to WGS84 at the centimeter-level, or in the local coordinate reference system. In the US, that would be NAD 1983 (2011). In Canada, it would be NAD 1983 CSRS. The precise version may be province-specific in Canada.

An offset of 3-5 ft / approx. 1 meter is classic for a difference between the earth-centered, earth-fixed ITRF (international terrestrial reference frame) and the plate-bound NAD 1983 frame.

When NAD 1983 and WGS 1984 were first defined, they were thought to be coincident although rumor has it that the ellipsoid calculations were done differently (I've heard on a cheap hand-held calculator by one participant) and so the GRS80 and WGS 1984 ellipsoids differ in the polar axis length. As it later turned out, NAD 1983 is not quite geocentric. Also, NAD 1983 is tied to the North American tectonic plate to try to minimize the changes to the coordinates of the control network (and everyone else's) over time.

1

This happend to me when exporting to State Plane NAD83 FT (w/ no .prj file).... Turns out my selected export units were "Feet", but I was viewing it in a projected coordinate system that assumed the units were "US Feet". Those extra decimal places really added up!

It was corrected by selecting "US Feet" as my export units in PFO.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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