GPS control

[Scott Tikalsky]

How were the original control targets set? Were they controlled by GPS and processed against State Plane coordinates? If so, Bob is right. If they were set with no calibration, then the coordinates may be referenced to the ellipsoid. If set conventionally, something else may be happening.

[Gromatici]

If you use GPS, have calibrated equipment, and properly post-process the data (to CCS83 with adjustments to the ellipsoid heights) then you'll have about .01' per 100' plus or minus difference between the Total Station set up and checking with your RTK system- uncalibrated (that's a different topic).

Your total station is shooting a horizontal line but there is a point scale factor and an elevation factor for each point on the grid surface so it not going to match without multiplying your measurement by those averages between the two points.
You may want to get "GPS for Land Surveyors" a read, and have a fellow surveyor sit down and explain these concepts to you.

Each point has it's own combination scale factor you can use to attain a "ground" set of coordinates. When staking in the field I usually threw is a project average in the Data Collector and then my shots to control points checked flat.

[Dave Lindell]

Every SPC coordinate is reported as it is on a grid surface, usually a plane with limited dimensions.

[Gromatici]

McGee said "Remembering that grid coordinates are the result of reducing local ground measuremnets to the ellipsoid surface and then projecting them onto a cone that slices the earth at the standard parallels for the zone."

I was wrong to say "at sea level". I was typing at 1am in the morning.

[McGee]

You may want to look at your calibration residuals, rotation and scale resulting from what is probably a two dimensional conformal transformation. The scale will give you an immediate sense of the integrity of the transformation solution. (Note, I take exception to calling a transformation a calibration because it connotes that all errors or differences are eliminated which is not the case). Now, if the 6 existing control points were SPC's then you would expect to get a scale factor equivalent to the combined factor for the SP zone, latitude and height above the NAD83 ellipsoid. Remembering that grid coordinates are the result of reducing local ground measuremnets to the ellipsoid surface and then projecting them onto a cone that slices the earth at the standard parallels for the zone. Otherwise, if the control is based on ground coordinates then you should get something close to 1.0 notwithstanding blunders. Again, what were your residuals ? There are some other things to look at but in any event I would expect that if there are no blunders in the 6 control points then your measured differences will be consistent in scale.

[McGee]

You may want to look at your calibration residuals, rotation and scale resulting from what is probably a two dimesional conformal transformation. The scale will give you an immediate sense of the integrity of the transformation solution. (Note, I take exception to calling a transformation a calibration because it connotes that all errors or differences are eliminated which is not the case). Now, if the 6 existing control points were SPC's then you would expect to get a scale factor equavalent to the combined factor for the SP zone, latitude and height above the NAD83 ellipsoid. Remembering that grid coordinates are the result of reducing local ground measuremnets to the ellipsoid surface and then projecting them onto a cone that slices the earth at the standard parallels for the zone. Otherwise, if the control is based on ground coordinates then you should get something close to 1.0 notwitstanding blunders. Again, what were your residuals ? There are some other things to look at but in any event I would expect that if there are no blunders in the 6 control points then your measured differences will be consistent in scale. Call me if you like.

[goodgps]

Robert,

I will assume that the H-V points were "GPS'd" using RTK methods.

RTK can sometimes "swirl" I call it the egg-shape swirl. There are many factors that cause this. You ma wish to read one of the publications suggested by the other folks.

If your epoch, is one of a higher RMS (root mean square) solution, an error will be introduced to your shot when it "fires" The shot can be right-on, or at the very edge of the egg-swirl. ie up or down .06' an out .03' or even more.

Remember, the RTK tolerance is +- 2cm vert and +- 1 cm horizontal. We like static for a better than 3rd order control survey.

Possible solutions: 1. the rodman must hold still.
2. use a tripod (even for rtk)
3. take multiple shots on the point then average.
note, this may not be totally correct, but heck !
4. if your rodman is "pole-ing" they can also rotate the direction they are facing and take 4 shots to mitigate rod errors such as bubble etc.

GPS requires great care. Never trust any GPS shot. Its a Great tool.
GOOD LUCK and may the PDOP be with you.

Dave