I'm working on a large tract in an area where I did quite a bit of surveying over several months of the Hot Season of 2011. The coordinate system of that project was NAD83(CORS96)Epoch 2002.0, obtained via OPUS, naturally.
Fast forward to the Less Hot Season of 2014 and I'm doing some more work at the tract. Naturally, OPUS still remains the method of choice for connecting to NAD83 and it only makes sense to publish coordinates at epoch 2010.0 that OPUS is presently returning. So, how to update the older coordinates?
The method that I hit on was just to resubmit all of the GPS observations on the eight OPUS points used in 2011 for new OPUS solutions in NAD83(2011)Epoch 2010.0 and readjust the works. The Precise Ephemerides that I hadn't waited for in 2011 were now, of course, available, and the coordinates of the CORS sites used had been updated to more accurate values, so the OPUS solutions looked startlingly good. Even the OPUS-RS solutions. The readjustment was the matter of a flick of a key in Star*Net.
One question that I had was whether the main control point, a Rod and Cap No. 10, had been as stable as I expected it would be during the three years between getting its position in 2011 and returning to do additional work. I had chosen the location for the control point with at least some thought to stability, but you never know for certain what you'll end up with until you do.
During the extreme drought that continued after August, 2011, the top of the aluminum cap appeared to rise a couple of inches above ground level as the top soil dried and shrank. The fact that the aluminum cap did not move with the surface of the soil was a good sign. It meant that I likely had set the roughly 30 in. rebar into what I thought was a gravel deposit below the soil. The odds were in favor of stability.
Then, in this wet year, the top soil expanded and rose to the level of the top of the cap again. In my experience, whenever you are seeing wet/dry vertical movements of a couple of inches or more, there are always reasonable questions about horizontal stability of a shallow mark in that soil.
So, yesterday, I collected a little over five hours of GPS observations on the monument and submitted it to OPUS for a solution.
The diagram below shows the situation. The points labeled "10a" and "10Day231" at the center of the larger error ellipse are the positions returned by OPUS from yesterday's work.
The other six positions to the left of the larger ellipse were from the work in 2011, with Pt. No. 10 being the least squares estimate of the coordinates of the point from those six OPUS solutions, 10Day215, 10Day220, and so on.
Since the 95% confidence error ellipse (drawn at the same scale as the point plot) estimated for the position returned by OPUS from yesterday's work actually takes in (if barely) the best estimate of Pt. 10 from 2011, there is no conclusive evidence of movement. Said another way: "The hypothesis that Pt. No. 10 has not moved is not disproven".
The distance from Pt. No. 10, the 2011 estimate of the position of the Rod and Cap monument to Pt. No. 10a, that from yesterday's OPUS solution, is 0.021 ft. horizontally, with a vertical difference of 0.016 ft. So we're not talking the End of Civilization if the mark has actually shifted a bit.
The work from 2011 when readjusted with fresh OPUS positions?
[pre]
Adjusted GPS Vector Observations (FeetUS)
From Component Adj Value Residual StdErr StdRes
To
10Day190 Delta-N 0.006 0.006 0.012 0.5
10 Delta-E -0.004 -0.004 0.006 0.6
Delta-U -0.039 -0.039 0.031 1.3
10Day195 Delta-N 0.012 0.016 0.012 0.9
10 Delta-E -0.000 -0.000 0.007 0.1
Delta-U -0.036 -0.036 0.032 1.1
10Day201 Delta-N -0.004 -0.004 0.012 0.3
10 Delta-E -0.007 -0.007 0.006 1.0
Delta-U -0.010 -0.010 0.031 0.3
10Day215 Delta-N -0.013 -0.013 0.012 1.0
10 Delta-E 0.000 0.000 0.006 0.1
Delta-U 0.040 0.040 0.033 1.2
10Day220 Delta-N -0.011 -0.011 0.016 0.6
10 Delta-E -0.006 -0.006 0.008 0.7
Delta-U 0.023 0.023 0.045 0.5
10Day221 Delta-N 0.018 0.018 0.015 1.2
10 Delta-E 0.000 0.000 0.007 0.1
Delta-U 0.007 0.007 0.039 0.2
[/pre]
I should have mentioned the durations of the occupations from which the above OPUS positions were computed. The GPS receiver was a base that was parked on Pt. 10 while secondary ties were surveyed from it, so the longish sessions on Pt. 10 and the associated OPUS solutions came along at very little additional cost.
[pre]
Day190 8.0 hrs.
Day195 6.0
Day210 7.1
Day215 5.7
Day220 4.5
Day221 5.9
[/pre]
HTDP is also a good tool for transforming epoch to epoch. The thing to remember is to set the input epoch at 2002.0 and not the date of the observation and for the output epoch to be 2010.0 and not today's date.
It would be interesting to see how the HTDP results compare to your resubmission to OPUS.
> It would be interesting to see how the HTDP results compare to your resubmission to OPUS.
Somewhat surprisingly, HTDP gave merely an approximation of the corrections that readjustment with the recomputed OPUS solutions gave. HTDP was not an excellent substitute for recomputation.
For example, take the case of Pt. 10, which was pretty well positioned both in relation to Epoch 2002.0 and to Epoch 2010.0 via multiple OPUS solutions.
Per OPUS solutions:
[pre]
Pt. N(ft) E(ft)
10 __3.451 __2.362 NAD83(CORS96)Epoch 2002.0
10 __3.478 __2.422 NAD83(2011)Epoch 2010.0
-----------------------
Delta +0.027 +0.060
{/pre]
Per HTDP predicted velocities at Pt. 10:
[pre]
0.66 mm/yr north = +0.017 ft. N from 2002.0 to 2010.0
1.32 mm/yr east = +0.034 ft. E from 2002.0 to 2010.0
[/pre]
Admittedly, an error of 0.026 ft. isn't much, but it's an error of close to 100%
> I should have mentioned the durations of the occupations from which the above OPUS positions were computed.
[pre]
Day190 8.0 hrs.
Day195 6.0Day210 7.1
Day201 7.1
Day215 5.7
Day220 4.5
Day221 5.9
[/pre]
A feature of the plot posted above is that the two most accurate OPUS solutions (in the sense of being closest to the least squares estimate from all six OPUS solutions), those of Day190 and Day201, were, as one would expect, also from significantly longer sessions.
Results like these are why I store data as opposed to coordinate files. It makes life simple in so many ways...
> Results like these are why I store data as opposed to coordinate files. It makes life simple in so many ways...
Yes, logging observations and adjusting the observations in the office really is long-lasting relief from sharp pains in the coordinate system.
