I used OPUS-DB back when it first came out, but haven't used it since.
A question-I submit a file, and it is accepted, so now there is a new station in the database. If I submit additional occupations, and give the PID from the first, what will it do with these other solutions? Does it just publish the first solution, or does it update the coordinates on the datasheet with a "mean" value (I imagine it would be weighted).
Howdy Mr Hamilton
Obviously the definitive answer comes from NGS. If you look at a point with multiple occupations (e.g. AH1762) you will see that current practice is to merely link to the most recent OPUS-DB solution with others linked from it. These solutions are in a separate database actually. Note that in these solutions whatever GEOID model was current at the time is shown and used to derive the NAVD 88 consistent height. The OPUS-DB solutions are static and not built from the data base like the regular datasheets are.
I do not know what the plans are for the future but the question you asked has not been answered as far as I can tell.
I am happy to see that NGS has tapped this resource for the latest hybrid model.
BTW, PID AH1762 is an interesting example of bad NGS policy with respect to superseding differentially determined heights. A good NAVD 88 height by leveling was superseded with a bad (decimeter level) GPS derived value. Note that the OPUS-DB ellipsoid heights and geoid ellipsoid separation more closely approximate the differentially determined value that the current GPS OBS height in the data sheet.
If you get some info from NGS on this matter I hope you'll post it.
Cheers,
DMM
Thanks, Mike. As I suspected. I knew there was talk about someday combining the various solutions on a station, and about readjusting when there are new realizations.
More reasons why bluebooking is superior (but still a PITA).
I guess if OPUS POJECTS ever gets released, it may resolve some of these issues (maybe).
to answer my own question...
OK, so I submitted one occupation from 2011. It was assigned a PID of BBCV43.
I then submitted another occupation, and told it that it was existing, and entered in that PID. Now the later submittal shows up for that PID, but it does have, under observed, a "See Also 2012-08-23" link.
So, as GeeOddMike says, it uses the latest one rather than somehow merging the results.
The whole reason I had to do this is that there is an NSRS benchmark nearby (that I submitted in 1989), HL 746=MR 81 RB (PID JW1808), just outside the fence. Each project must have an NSRS mark with a NAVD88 elevation. The HL746 mark, in order to qualify, must be checked against another mark. The next two BM's upstream and the next two downstream, along the RR, are gone. But, the USACE will accept a OPUS-DB elevation. The OPUS-DB misses our second order tie with invar rod to HL 746 by 0.015 m.
Someday, composite solutions
From an NGSer,
We are starting to work on this, as quite a few marks now have multiple published OPUS solutions.
In the interim,
yes, the individual solutions merely stack up on top of each other, leaving the user to manually hash them together or pick their favorite.
Soon,
One data sheet will provide the "best" position and accuracy, with a "details" link to scatter plots, time series, and individual solutions. Before publishing a new recovery, you should see how it stacks up against previous uploads as a closure check of sorts.
What is "best"?
- For position, probably an average, weighted to preference the "better" solutions (newer, longer duration, lower peak-to-peak, more coherent).
- For accuracy, some amalgam of precision, peak-to-peak, and rules of thumb, with any mark lacking repeated or recent observation fading as "NO CHECK" over time.
Compiling the OPUS solutions is pretty easy, but adding -projects and -NGSIDB results is slightly trickier, as each is a bit different. We hope the result will be as simple as the current data sheet, while offering more background for the detail-oriented.
PS: NAD83(2011) isn't shown on OPUS data sheets yet, but will soon, when some GEOID and plumbing issues get straightened out, we will simply re-solve everything from the original data files and rebrand the existing solutions as "superseded."
Thanks for posting and sharing opinions about OPUS. It is great to have an open forum to get feedback and work out user ideas and concerns.
Someday, composite solutions
Joe: Thanks for the NGS insight. I look forward to seeing these enhancements. I see this is your first post (and first day registered). I hope you will maintain a presence here, we always could use more of the viewpoint from "the other side"
