A later thread on vertically mounted or sky-obscured marks
When you say 'low using geoid xxx', what us the method of deriving the ellupsoud height??ÿ
I will assume he means what OPUS returns, pretty much what I would expect to see; 4-5cm
Yes, I was using the OPUS report numbers.?ÿ The main report gives ellipsoidal height as measured by the session, and the ortho NAVD88 value assuming Geoid12b.?ÿ They also tack on at the bottom the NAVD88 calculated using xGeoid17b.
The sessions I've done nearer to points used to define Geoid12b have usually fallen within a cm or two of the leveled NAVD88 value.?ÿ The sessions differing by 3 to 5 cm were on marks at considerably more distance from those defining points.?ÿ There is of course some offset over the general area between the geoid models, but I found consistency over a much larger area in the differences I got using xGeoid17b.?ÿ That makes me think my sessions were valid and a confirmation of the improvement 17b or later geoids will bring.
Do you have enough equipment to obtain some simultaneous data? I often find OPUS to provide solutions well outside 3cm, even when quality undicators say it's 'better'.
It is important to note that the xGEOID models (currently xGEOID17 and later this year xGEOID18) do not provide an estimate of NAVD 88.?ÿ The model computes the expected changes in the geometric position (lat,long,ellip hgt) relative to the North American Terrestrial Reference Frame of 2022 (NATRF 2022) and heights in the North American-Pacific Geopotential Datum of 2022 (NAPGD 2022).?ÿ These values should be used for planning purposes for surveyors and their clients to consider the impact of the datum changes on the array of their products and services relative to current NAD 83/NAVD 88.
Yes, I mis-spoke when I linked NAVD88 with xGeoid17b.?ÿ Just read that as orthometric assuming xGeoid17b.?ÿ That definition of 17b is why the offset I mentioned exists.
BUT, there is a use for the experimental models. I have had several projects recently where I suspect the GEOID12B model is tilted, but I am not sure if it is that or if there is a problem in the NAVD88 leveling. I use the geoid model in a RELATIVE mode in adjustment rather than in an absolute. I hold one or more orthometric heights, and the adjustment will apply the differences in geoid height between points. If that improves the results versus the published NAVD88 orthometric heights, then I know the problem is with the geoid model and not the published heights.?ÿ
Here is an example. We did a linear project along the Schuylkill Expressway near Philadelphia. Occupied seven NSRS benchmarks. There was an apparent tilt in the misclosures. So I did a comparison using GEOID12B and xGEOID16 (this was in January 2017) on the two end benchmarks.?ÿ?ÿ
So xGEOID16 improved the misclosure from 5.3 cm to 2.2 cm.?ÿ?ÿ
?ÿ
Do you have enough equipment to obtain some simultaneous data? I often find OPUS to provide solutions well outside 3cm, even when quality undicators say it's 'better'.
I only have one receiver.?ÿ
My limited experience indicates in my area 4-6 hour sessions under good conditions usually get results repeatable within the ellip ht pk-pk range.
The OPUS reported ortho range is unrealistically optimistic?ÿ and has no valid theoretical basis because it is not tied to the geoid uncertainty in the area.
John's example is essentially using the xGeoid the same way I was above.