We normally use OPUS to get coordinates for our base. We always deliver our data in state plane coordinates with NAVD88 elevations. OPUS makes this very easy as all I have to do is convert the SPC from meters to US survey feet.?ÿ
Anyway since OPUS is down I have been looking into using Trimble RTX. As a test I submitted a RINEX file from last month so I could compare and get comfortable with the results before I relied on this service for our current project.?ÿ
My problem seems to be with converting ellipsoid heights to orthometric elevations. Below are screen shots from an OPUS solution?ÿ& RTX for the same observation file.?ÿ
The X,Y,Z are close enough between the two for my needs but I can't seem to get a good conversion from the ellipsoid height to a NAVD88 elevation. I tried using UNAVCO (which I have never used before) and I get this using the data just from OPUS. It reports the elevation as 401.6 meters vs OPUS 400.348 a difference of over 4 feet. I assume the problem is OPUS uses GEOID 12B and UNAVCO is using EGM96 but I am not positive on this.?ÿ
Can anyone help me learn how to get solutions that will jive with OPUS? Our work is not super precision critical. Contractually it is specified as submeter but I think with our normal methodology we are much better than that.?ÿ
?ÿ
?ÿ
Mark Silver posted links to GEOID12b a few weeks back, you should be able to access it here:
?ÿ
I know nothing about Trimble RTX. Looking at the screen capture above reveals that it transforms the ellipsoid height via the Earth Gravity Model 96 (EGM96). As noted this is related to ??mean sea level.?
Unfortunately, heights compatible with NAVD88 are computed from NAD83 ellipsoid heights and a HYBRID geoid model (currently GEOID12B). The hybrid geoid model is created from a gravimetric model plus datum transformations (since NAD83 is not geocentric) and a corrector surface (from the GPS on benchmarks data set).
Note that NAD83 and IGS/ITRF ellipsoid heights differ in your solutions at the > one meter level. This shows the importance of knowing what ellipsoid heights are to be used with which model of ellipsoid-geoid separation to use.
Loyal helpfully provides a link to an alternate source of the data and utility to compute the separation. I did note that Mr. Silver??s site provides GEOID12A and not GEOID12B data files. The ??B? model incorporates corrections to the ??A? model impacting limited geographic areas including the Dallas/Ft Worth, TX area.
BTW, EGM96 is not even the most recent model. See:?ÿ http://earth-info.nga.mil/GandG/wgs84/gravitymod/egm2008/egm08_wgs84.html
Are you using Trimble GPS?
Yes - I am using R10.
OPUS appears to be back up, as is most of the NGS website. Not sure for how long though?
https://geodesy.noaa.gov/OPUS/
I use TBC and Access for the R-10. If you use those then you can set up a projection file with Geoid12 and enter the lat, long, height and it will convert the height to a NAVD88 elevation based on Geoid12. If you don't have Geoid12 it's available on the Trimble website to download.?ÿ
Yes I saw OPUS was back up. I just submitted the files. Have not got them back yet.?ÿ
I don't have TBC. I do pretty much everything in the data collector or Excel.?ÿ
?ÿ
I do pretty much everything in the data collector or Excel.?ÿ
?ÿ
I use Survey Pro, and in there I have Geoid12B loaded. So I could use the data collector to apply the appropriate geoid height to the ellipsoid height if I needed to. I don't, because I use StarNet to do that, when it's necessary.?ÿ
You could use StarNet to do this as well. The demo version will do it, up to 10 points at a time after the 10 day trial. You would have to load the appropriate geoid data file.?ÿ
This brings up the matter of the propriety of relying solely on your data collector for data resolution.?ÿ ?ÿ
?ÿ
I do pretty much everything in the data collector or Excel.?ÿ
?ÿ
I use Survey Pro, and in there I have Geoid12B loaded. So I could use the data collector to apply the appropriate geoid height to the ellipsoid height if I needed to. I don't, because I use StarNet to do that, when it's necessary.?ÿ
You could use StarNet to do this as well. The demo version will do it, up to 10 points at a time after the 10 day trial. You would have to load the appropriate geoid data file.?ÿ
This brings up the matter of the propriety of relying solely on your data collector for data resolution.?ÿ ?ÿ
?ÿ
The DC should handle a geoid model.?ÿ
The DC should handle a geoid model.?ÿ
It does and I have it loaded but I am not sure how to manually convert heights like that. I think it uses the good model for its internal calculations on the fly??ÿ
?ÿ
It's hands free.
I just did this for a point the last time I was in the field, select key in point, select WGS84 (I know it should say NAD83 but it doesn't, yours may correctly have NAD83 as an option), key in name, lat long height, feature code, then?ÿreview point, select grid and it displays the XYZ. This was with my R8 system and using survey controller on a TSC2.
It does and I have it loaded but I am not sure how to manually convert heights like that. I think it uses the good model for its internal calculations on the fly??ÿ
?ÿ
In Survey Pro it is done in the point entry screen. You have the option to enter points as Lat/Long/Ellipsoid Ht or as grid coordinates with ortho elevation.?ÿ Once you've entered one you can flip the display to the other.
GeeOddMike:
RTX does not use ANY geoid model, it only supplies an ellipsoidal height. It computes in ITRF current epoch, then transforms to NAD83 (2011) epoch 2010.0 (if selected).
As to the original poster, the difference in ellipsoidal height between OPUS and RTX is only 14 mm, which I think is very good.?ÿ
GeeOddMike:
RTX does not use ANY geoid model, it only supplies an ellipsoidal height. It computes in ITRF current epoch, then transforms to NAD83 (2011) epoch 2010.0 (if selected).
As to the original poster, the difference in ellipsoidal height between OPUS and RTX is only 14 mm, which I think is very good.?ÿ
Thanks. Yes I think the RTX OPUS comparison is good. I just was having problems converting ellipsoidal heights to elevations.?ÿ
OPUS was back up again and I got my solutions and sent my data off today. I have an 11 hour drive tomorrow to the next job but will take some time and play around with the data collector and/or the programs Loyal linked to that Mark Silver posted and see if I can do the conversions. My problem is temporarily solved but I would like to learn how to do the calculations in the future. Even if OPUS is running the RTX seems to provide much faster results.?ÿ
?ÿ
If you look at the RTX output posted you will see a height indicated to be orthometric and determined using EGM96. The original poster reports it differs by ??...over 4 feet..?
I agree that there is good agreement between the two tools as far as in the computed NAD83 ellipsoid height.
Whether some users benefit from the EGM96 based heights, those in the US expecting to get good agreement with NAVD88 will be disappointed.