NAVD 88 Elevations

Should work, but it's only going to be as accurate as the GPS constellation is, or if you are working at a bad time, you might not have repeatable results.

What I would probably do is place the base at the site and a rover at the BM and run a static session. Then, process your base against OPUS and compare the two. From there I think you'll have a better idea of the transfer of the z component to the site, which as we all know, is the weakest link in the mix. I suspect that the static will yield a superior and verifiable result with regards to this. That being said, if they're more than a tenth different, something else is wrong.

The issue I have with OPUS-RS is that it's finicky at times and you can't really tell. If you use my method, you can then, not only process to your base, which is constrained and verified by OPUS, but process to any number of CORS stations around and see those results as well.

The static observation, at the BM, based on your distance, should be no less than 16 minutes, but I like to double that. If you did two different static shots on it, from the base, under different constellations, then you've tightened it up about as much as you can unless you tie in enough BM's surrounding the site to control it. Otherwise, just use one.

Set on the bench mark, if you can't then level a point and set on it, apply the newest geoid to the NAVD88 elevation then check into the bench mark system.

I have just over a tenth between an OPUS solution and NAVD88 which is a great improvement from previous EPOCHs........

But if your FIRM declares NAVD88 and lists the bench marks in the area as controlling (this is what I've seen), then it's what you have to use.

By applying the geoid to the NAVD88 ELEVATIONS you will raise or lower the ellipsoid height from an OPUS solution.

There are areas where the bench mark system isn't stable, so be careful...........

From a pure yes or no standpoint, I'd be pretty satisfied with two 3 minute RTK observations taken four hours apart IF they gave me two answers that agreed at a level I could live with. However, that would be the absolute bare minimum, and all the advice given above is good. I'd definitely run static on the base and correct it with both my own software and OPUS. And I completely agree with Kris that OPUS-RS is finicky at best, although it seems to be a lot better now than when it was introduced.

Base at unknown site with good sky view implies everything else has poor sky view. Will not comment on that.

Nor knowing where DB is surveying, OPUS-RS may or may not be an option.

1/ My preference is OPUS-RS for elevations. As soon as one says elevations , occupation times should be double.

2/ As soon as one says poor visibility, occupation times should be doubled.

3/ Two separate observations do not always prove good control, three give you a chance if one is an outlier.

OPUS-RS is always finicky about poor data. If you have a chance for poor data then you need to be able to switch to OPUS for what you can get. Considering 1, 2 & 3, you should start with a 4-8 hour static at minimum.

Better option is consider GNSS capable receivers.

Paul in PA

Just finally getting time to reply..

Kris Morgan - I'm wondering where you found your info on the relationship of baseline distances and occupation times (you stated occupation time at BM should be no less than 16 minutes with a 3-4 mile baseline).

All site conditions being equal, is there any accuracy gained by having the base set up at the site rather than on the roof of where I work, where it will be safe all day long, (occupation time will be long enough to use OPUS rather than OPUS-RS) while I perform 2 RTK (3 minutes or so) and static occupations (2x16 minutes??) separated by 4 hours or so at both the BM and the site? The sky view will be superior on the roof top.

I'd like to compare my own post-processed static data, the raw RTK data, the RTK data adjusted to the BM datum (if significantly different), and the OPUS solution.

I am not sure where your project is in California, but most of the CSRC Height Mod Projects were done in areas subject to subsidence, except maybe the Central Coast project, which may be what you are using based on the date. Regardless, I would take a close look at the time series plots of any PBO or NGS CORS stations in your project area, specifically the height component, before deciding to hold a NAVD88 elevation on a passive mark that was established 8 years ago. Especially if you are going to rely on only one NAVD88 mark to "shift" your project to that datum.

Add another 88 bench mark and check that your methodology works. Jp

The BM is part of the CSRC Central Coast Height Modernization project in Santa Rosa area.

What is the station name, or even better PID if it is in the NSRS database?

Does this height mod station show an accuracy greater that a tenth? Or is there a more precise local elevation?

Per PRC 8897, one field observed independent connection meets the requirement by the State to establish a COH88 value. Nothing in there states the duration and if you do not state the accuracy, you meet requirements.

However, I would also be hesitant to rely on only one passive mark and two short (1-3 minutes) measurements and would follow what NGS NOS 59 memorandum states. Plus, in my opinion only, the short observations wouldn't satisfy Item C2 per the elevation cert.

Looking at time series plots of PBO stations in that region, the vertical component looks to be pretty stable over time, other than a seasonal fluctuation of about +/- 2cm. However, I agree with Jp7191 that using two NAVD88 bench marks (height mod derived) would be better, depending on the what the ultimate goal is for the project.