Just FYI: Initially accidentally posted this to the wrong forum. So I'm just re-posting here. Anyway...
So I'm relatively new to all this. I am not a licensed surveyor. I'm a geologist with an engineering firm. And over the past couple years, I've been using a TOPCON GR3 when we need relatively accurate locations for our own internal projects. For example, we sometimes need to track the locations of groundwater treatment injection points over an active construction site. The survey nails and whiskers marking these locations get buried. Sometimes even monitoring wells get buried. So we've used the GR3 to establish our own relative field control points and then re-locate features over time. And it's been beyond awesome to see how well this equipment works-- awesome enough that I've been trying to read up more on GPS and geodesy and find ways to experiment with this equipment in the field when possible.?ÿ
I've recently been on projects near various benchmarks associated with bridges and such. So I've had a lot of fun running static sessions on these benchmarks. I use OPUS for post-processing. And the resulting coordinates have generally been within an inch of benchmark coordinates. Just this last week, however, I ran two separate 4+ hour static sessions on a benchmark with coordinates in NAD83 (1991.35). Our GR3 runs only in NAD83(2011). My understanding is that these coordinates can be transformed to NAD83(2011) with NGS NCAT using NAD83(HARN). And so I transformed the coordinates from NAD83(1991.35) to NAD83(2011) to compare against my OPUS solutions. My OPUS solution x,y are each about 5 inches off of the transformed benchmark coordinates. (The two OPUS solution x,y coordinates are about half an inch apart. Overall RMS is 0.016 meters for both, with peak x,y errors of about 0.008 to 0.013 meters. In the shorter session, 93% of observations were used. In the longer session, 88% of the observations were used.) I'm just curious: Is it common to see a horizontal difference of about 5 inches between an OPUS solution and coordinates associated with an older benchmark in an older datum realization??ÿ
I'm also curious about using older datum realizations with our GR3. The GR3 currently handles only NAD83(2011). In the case above, I wanted to reference our locations to this benchmark. So I transformed the benchmark NAD83(1991.35) coordinates to NAD83(2011) and started the base station with those transformed coordinates-- I ran the static sessions just because I was curious. My understanding is that there are probably small non-linear differences between different datum realizations, so if I wanted to run the GR3 with the NAD83(1991.35) coordinates, I should install that datum realization on our handheld FG-200, even though my understanding is that over short distances it almost certainly wouldn't matter. On the flip side, my understanding is that, technically, if I wanted to convert these coordinates (the ones acquired relative to the transformed benchmark) back to NAD83(1991.35), I should transform each separately due to potential non-linear differences, rather than calculate the offset at the benchmark and then apply it to each location-- even though, again, over a short distance, it almost certainly wouldn't matter? Does that make sense?
I've had a lot of fun reading these forums. And reading up generally on geodesy. It's fascinating stuff. And I apologize if some of what I wrote above is confusing or just embarrassingly wrong. I'm not a surveyor. But I seriously appreciate the opportunity here to pick so many brains and hopefully learn a bit more.?ÿ
Thanks!
We might be better able to comment if we had the PID to identify the one with the 5 inch discrepancy. The shift between datum realizations will be different around the country.
1991 was a long time ago and satellite availability and receivers have improved, too.
The GR3 only runs in NAD83(2011)? That doesn't seem correct to me, but then I've never used them.
First - a "benchmark" is generally an elevation reference. You seem to be concerned with horizontal coordinates. Lets call this a "Geodetic Control Monument"
Second - Your GR3 is not fundamentally restricted to NAD83(2011). The way you use it may yield results in NAD83(2011). But that is in your process, not the receiver.
Third - As Bill said, it would be great to see the datasheet for this monument. Just because it shiny & brass doesn't mean it's coordinates are any good. Not all control is created equal. For example, if it's a NGS monument that was established by "classical geodetic methods" you will be lucky to agree with it by a couple feet.?ÿ
Fourth - The difference between NAD83(91) and NAD83(2011) in my area is about 5 inches. So I'm suspecting some problem with your transformation. These utilities the NGS puts out are rubber sheet estimators. Not 100% reliable. The way I do it when I have to is to get some datasheets for monuments in the neighborhood which have values in both datums and figure out the shift, then apply that to my monument.?ÿ ?ÿ ?ÿ
Fifth - there could be a multipath bias in your OPUS positions, but not likely.
Sixth - isn't geology enough of a challenge? My creditors children are cold and hungry.?ÿ?ÿ
Thanks for the responses all! Few things:
1. This isn't an NGS monument. It's a Caltrans "benchmark" for a bridge. It's not depicted by the NGS Data Explorer ( https://www.ngs.noaa.gov/NGSDataExplorer/ ). I'm using the term "benchmark" a bit loosely, as it was described to me in an email. The as-constructed drawings refer to it as "Survey Control."
2. I'm sure the GR3 can run other systems/datums. But when selecting the coordinate system in the FG-200, only NAD83(2011) and some other variant are available. I'm assuming I would have to somehow install additional systems/datums. My understanding is that this equipment was bought used. I inherited it from a previous GIS fellow who had helped acquire it through his connections. Still, my interest in using past NAD83(2011) realizations with the GR3 isn't so much about wanting to actually do it. I don't. At least, not now. I just want to see if I understand the theoretical consequences of various datum realizations and how they would impact using this instrument in various scenarios.?ÿ
3. Geology is pretty rad. But I'm becoming obsessed with this crazy notion that we can somehow assign a coordinate to any given point on the surface of Earth and then someone manage to actually relocate it later. The whole concept is wild.
You aren't doing what you think you're doing. It would be like saying the pencil and field book you're running a level line with only works with NGVD29 elevations, and can't work with NAVD88 elevations
Thanks for the responses all! Few things:
1. This isn't an NGS monument. It's a Caltrans "benchmark" for a bridge. It's not depicted by the NGS Data Explorer ( https://www.ngs.noaa.gov/NGSDataExplorer/ ). I'm using the term "benchmark" a bit loosely, as it was described to me in an email. The as-constructed drawings refer to it as "Survey Control."
It could be that it is a benchmark and the horizontal coordinates were only loosely acquired?ÿ for the purpose of spotting the thing on a map. Long story short - it's likely that your OPUS coordinates are the ones to go with.?ÿ
The ability of on-line GNSS processing tools to replicate published positions for survey monuments is dependent on the quality of the position we seek to replicate, the quality of our GNSS data, and attention to geophysical changes in the time interval between the original determination and our work.?ÿ
As you use the US NGS OPUS processor you are able to obtain NAD83(2011) coordinates. There are alternative services that do not provide NAD coordinates. Remember the disclaimer on the OPUS solutions that the position provided is only valid if you occupied the correct point (not its reference mark of a point with a similar name, etc), you chose the correct antenna model and measured your height above the monument correctly. Also consider that something localized could have impacted the survey monument: frost heave, displacement due to erosion of other impacts.
As you are working in California, and as a geologist, you undoubtedly are aware of changes due to factors like plate motion and subsidence. Can a survey monument be said to be in the same place on the surface of the earth in an area undergoing these phenomena? Over shorter distances it is possible that the changes are correlated. As distances from CORS sites can frequently reach tens and hundreds of kilometers, we cannot make that assumption.
The position in an OPUS solution is provided in the current NAD83(2011) system and at the date/time observations were made. The position is determined with respect to the CORS network. Examining the CORS position file will show the coordinates and velocities at the NAD83(2011) reference epoch. Velocities are expressed in both dX, dYm, dZ and dN, dE, and dU. ?ÿNote in the attached file that coordinate file for TORP shows annual changes of -1.4, 2.9 and 2.4 centimeters in X,Y and Z or 3.1, -2.6 and -0.2 centimeters in North, East and Up.?ÿ
The NGS has a tool, HTDP, allowing a coordinate to be transformed between times. It can also be used to transform between reference frames.
Recommended reading on HTDP:
https://www.ngs.noaa.gov/TOOLS/Htdp/Htdp.shtml
https://www.ngs.noaa.gov/TOOLS/Htdp/HTDP-user-guide.pdf
https://www.ngs.noaa.gov/TOOLS/Htdp/Pearson_Snay_2012.pdf
This long-winded reply to the original post is only intended to highlight the fact that a well-determined position at a specific date can be updated for comparison through transformation tools.
If the position was initially determined by classical (optical) geodetic methods we have the unresolvable problem (unless we have the original observations and the point can be attached to the NAD83(2011) network) that the point was never included in the post1986 GPS-only adjustments.
I applaud your pursuit of knowledge on the issues related to validating published coordinates. Should you wish to pursue you quest further, I suggest contacting the NGS Advisor whose contact information is attached below.
Real world coordinates aren't going to be obtained by translation. Caltrans probably ran control back in the 90's by occupying nearby HARN points. To correctly recreate the coordinates it would be best to reoccupy the HARN points and the Caltrans monument. Then process the data files. Assuming the HARN points still exist, they haven't drifted away from each other and the monument has moved along with them you should get very close to the same result. But the translation programs ,,,,,,never seen them work for that.?ÿ
The important issue is that if you are giving out data to clients for a project with established control, use it. If all you are doing is your own thing then go ahead and put 2011 numbers on points.
?ÿ
NAD83(2011) is already 7 years old, it is a snapshot in time and is degrading as CORS points are moving away from each other.
"Is it common to see a horizontal difference of about 5 inches between an OPUS solution and coordinates associated with an older benchmark in an older datum realization?"
Yes, in California it is. Much more than that in some locations. Here is an HTDP modeled solution from 1991.35 to NAD83(2011)2010.00 for a point near Oakland California (PID AA3814). Yes, that is 1.56' (0.47m) by 1.06' (0.32m)
Also you mentioned transforming from HARN - 1991 predates the HARN. In my area HARN and 2011 agree very well.
The epoch date for the California High Precision Geodetic Network (HPGN), which was our naming convention for the HARN campaign, is 1991.35. More correctly, NAD83(1992) Epoch 1991.35.?ÿ
There is just over a tenth of a foot for the published 83/93 (HARN) value for the?ÿlocal HARN point and the 83/2011 value, mostly in the Latitude. This of course will not exactly match an OPUS number, but from what we hear about California, their CORS points are sliding really quickly, I guess heading west towards the ocean.
