Benchmarks vs GPS cors

[laveyb24]

Figured I would post this question to the community and see what everyone thinks.

I have done a small static network in Southern California. I collected 10 NGS benchmarks that are first order class II. When i process the data i am noticing the 10 NGS benchmarks which referance geiod 12b all seem to work with one another. The cors stations all seem to work with each other as well but the 2 do not agree with one another. Both are NAVD88 and referance geiod 12b. I would like the network to agree with the CRTN network. Which would you hold and why? Thanks in advance.

[William Magee]

On the NGS datasheets, what method does it state for the determination of the orthometric height?

Do they state it was by differential leveling, or by GPS observation and high resolution geoid model, or otherwise?

[laveyb24]

Differential Leveling.

[TTaylor]

2 quick questions:
1.) What is the magnitude of the difference, and
2.) What is the date of the surveys that produced the NGS values and the CRTN values?

[Steve Martin]

The published geoid height is not necessarily the geoid height that was used to derive the NAVD88 height published for a CORS station.

Given the choice between a 1st or 2nd order leveled benchmark elevation that was part of the definition of NAVD88, and an NAVD88 height derived from some subsequent survey using an unknown-to-you geoid model, I would choose the leveled ADJUSTED NAVD88 elevations over the derived heights on the CORS stations. Just verify that that is indeed what you have first.

Also note that the CRTN epoch 2011 is based upon the older NSRS 2007.00 datum and GEOID12B is optimized for the newer NGS NAD83 (2011) epoch 2010.00 datum.

[William Magee]

Steve has some very succinct and sound advice.

[David Kendall]

Thanks for the discussion, this is all good stuff for me as I have recently been trying to reconcile the differences between static GPS observations derived through OPUS (CORS) with private RTN based RTK observations (CORS? CRTN? call the company and ask them?) as well as with published benchmark data. It's enough to make my head spin.

Our orthometric height variations are 1.5 - 3 feet across several different orders of benchmarks, similar to OP and we are working in SE Sonoma County.

Just enough to make me wonder on a flood certificate if I'm doing the right thing....

[SPMPLS]

Steve did an excellent job of identifying the "apples" and "oranges" involved, likely contributing to the height differences the OP is seeing. One must understand how hybrid geoid height models, like 12B, are derived and which realization of NAD83 they are to be paired with. No orthometric heights were changed as a result of the NAD83(2011) 2010.00 realization, meaning the geoid and ellipsoid heights changed.

As to SE Sonoma County, there is zero NAVD88 leveling in that area, so obtaining NAVD88 ortho heights is entirely dependent on CORS stations and hybrid geoid models.

As an aside, an epoch 2017.00 update of the California Spatial Reference Network and CRTN is being completed right now by the CSRC and will be out late this summer.

[Steve Martin]

Thanks Scott,

We use Geoid09 with the CRTN epoch 2011 to follow NGS's recommendations.

If we do have a project that does not need vertical control, such as a boundary survey, I'll call the derived NAVD88 heights "Nominal NAVD88 elevations" since they are not tied to an actual benchmark.