It depends. Sound familiar.
For the most part, we run with what we can find that seems to have been determined and used previously for serious purposes. Sometimes those are the RM's from years back. Sometimes those are highway or bridge related solid points. Sometimes we go with city-provided data for manholes and such. Always tie to more than one to avoid blunders. And, anymore, convert the old numbers to the new numbers. That is an increase to the old number of about 0.45 feet in most of the area we cover.
(Knock on wood) We have never had an elevation certificate questioned because of the reference benchmark cited.
I use the bench marks shown on the panels, I would never use OPUS in my area because it doesn't work. The bench marks were leveled and if you level between them now they still hold the relationship that is published, proving that they are stable or the entire area has moved up or down the same amount. Unlikely.
OPUS and/or CORS using a geoid model Will not return elevations matching the benchmarks.
Depending on the geoid
Model of your choice you will be high between .2 to .4'. Not bad for a rough number, but not acceptable when you want real numbers. Been told other areas have better results, but not so much here. The important thing about a geoid model is does it "model" the geoid slope?
Once you see that it does then it's simple to get real numbers from GPS.
The way to do that is obvious.
My advice is to do extensive testing of your equipment and conditions in your area, remember, it's not as important that the model return correct elevations when an OPUS number is surveyed but rather that the model returns the correct difference between benchmarks. If it does that then you have a model that is "modeling" the area correctly and from there it's easy to get good data.
$1000
More if the building is complicated.
>you STILL have to then compute elevation based on the current geoid.
If they want NAVD88 elevations, what does the current geoid have to do with it?
Bill,
"If they want NAVD88 elevations, what does the current geoid have to do with it?"
Good question. Now I have to say that my shallow and pitiable understanding of vertical datums has been shaken to the core. Is the current geoid a "flavor" of NAVD88, or something else entirely?
Dave
Cliff,
Day-um! This guy, Gunter Hein, is the Master of Dense. Here's how he starts off:
His title, all by itself, is breathtaking:
"Orthometric Height Determination Using GPS Observations and the Integrated Geodesy Adjustment Model"
Dave
The closest one I can find. B-)
:good: 😀
County?
The NGS publishes ortho heights on their data sheets. I level from of their monuments and back again. What elevation do I have on my site? Is THAT acceptable?
County?
Absolutely yes! FEMA and the Federal Justice Department have held that any professional that relies on current data held in the National Spatial Reference System of the National Geodetic Survey is properly referencing sites for coverage in the National Flood Insurance Program.
Some areas of the United States are known crustal motion areas, and if you practice there, the absolute minimum number of benchmarks in any loop is two. Note that NGS itself ties to a minimum of THREE benchmarks at both ends of a leveling section.
Geoids are matched with the appropriate reference frame adjustment. You are not supposed to mix and match. It angers the Gods.
Cliff,
Correct me if I'm wrong, but isn't NAVD88 an ellipsoid? And isn't Geoid 12A a bumpy, irregular shape? Where does the reference frame fit into this? How do you "appropriately" match all three up?
I am loath to anger the Deities. 
Dave
upheaval
Where would I find information on Somerset County, Pa and it's up and down motion or lack thereof? Is this info on the FIRM? Is there a website outlining areas of the US that physically undulate? I've been looking around for that info for an "Elevation Form."
Sorry, Cliff, I'm not sure how to interpret that. Could you explain again in a little more detail?
Are you saying that NAVD88 has flavors that match the NAD83 flavors? Have the NAVD88 elevations changed with each horizontal adjustment?
Bill - Yes, the geoid models are designed to best fit different realizations of NAD83; GEOID12A is designed to fit NAD83 (2011). It is not appropriate to use older geoid models with the 2011 datum.
Dave NAVD88 is an approximation of mean sea level; I'm sure Cliff can explain it better than I can. MSL is an undulating surface that changes with changes in gravity due to (changes in) sub-surface density.
NAD83 is an ellipsoidal datum that is based on the GRS80 ellipsoid. GRS80 and WGS84 are essentially identical in size and shape but differ by about 2 meters at the origin.
NGS computes its geoids based on the current realization of NAD83. The recent horizontal re-adjustment of NAD83 (2011) is the current realization of the horizontal datum based on observations of the National network of CORS. It is that reference frame that is the horizontal basis of GEOID12 and its addendum, GEOID12a. The prior realization NAD83(1996) was the reference frame for earlier editions of the geoid, including GEOID96, a special recompilation of the EGM96 geoid that was world-wide. NGS publishes 7-parameter transformation constants to move from reference frame to reference frame for the entire country. When NAD83 (2011) came out a few years ago, I computed the 7-parameter transformation from 1996 to 2011 based on just our CORS sites in Louisiana and presented it to the Louisiana Society of Professional Surveyors annual convention in Marksville, LA.
All datums evolve based on new observations. NAD83 will likely continue to evolve through new realizations until the legendary new datum is published sometime around/after 2022. That will be quite a spectacular event as it will be combined with all of the data being collected through GRAV-D, and it will be the FIRST CONTINENTAL THREE-DIMENSIONAL GEODETIC DATUM IN HISTORY.
Louisiana is in an area of crustal motion. We are working on a state-wide Quasi-geoid that will provide continuing coverage and accuracy as the area continues to subside. We are doing it with terrestrial gravity measurements, specifically ultra-precise relative gravity at benchmarks located in-between our state-wide network of LSU CORS sites that have absolute gravity observations, some with multiple absolute observations.
The objective for all of this: make life simpler for the professional practitioner.
upheaval
If an area is subject to crustal motion, the NGS Data Sheet will advise to that effect. (That includes areas subject to UPLIFT, such as isostatic rebound from the last Ice Age.)
Cliff,
"The objective for all of this: make life simpler for the professional practitioner."
Is it just me, or is that objective headed in a contrary direction?
Dave
Well, some people DO wear steel-toed boots at the pistol range ...
When computed from a theoretical geoid, yes. There's only one ellipsoid, but that's theoretical. There's only one geoid, but that's theoretical. Although the ellipsoid won't change, the scientific guesses where the geoid is with respect to the ellipsoid are an iterative problem that supposedly is converging towards the truth. The more the observations, the better the guess.
