Notifications
Clear all

NGS Benchmarks versus OPUS derived Elevations

55 Posts
21 Users
0 Reactions
54 Views
(@skeeter1996)
Posts: 1333
Member
Topic starter
 

I'm currently in a dispute with a Community Flood Plain Coordinator over the basis of determining elevations for FEMA Elevation Certificate's. I regularly observe one foot differences in OPUS derived elevations when checking in to NGS monuments with published elevations. I can see what elements are involved with OPUS derived elevations, but not NGS monument elevations. Therefore I tend to favor the OPUS Elevations.

The dispute with the Community Official came about because the installation of a sewer system nearby has elevations derived from a NGS benchmark and my OPUS derived elevations (two separate observations which differed .003 meters). The OPUS derived elevations result have the building site over a foot above the BFE. The NGS monument result are the building site is 0.1 foot below the BFE.

I realize this difference in distance should be considered insignificant, but try explaining that to a County official. I'm interested in others opinions.

 
Posted : August 4, 2022 8:18 am
andy-j
(@andy-j)
Posts: 3121
Member
 

Interesting situation.?ÿ What happens when all the actual monuments are gone and not getting replaced??ÿ ?ÿCan you share where you are/what monuments you are comparing to??ÿ I wonder if others have similar results in your area.?ÿ?ÿ

 
Posted : August 4, 2022 8:25 am
OleManRiver
(@olemanriver)
Posts: 2459
Member
 

When you say 1 foot from OPUS vs a NGS Benchmark. The first thing that comes to mind is NAVD88 vs NGVD29. Roughly. ?ÿWhen I compare OPUS to a NGS Benchmark I donƒ??t see that much difference unless 1. Disturbed monument 2. Multi path environment 3. I chose the wrong datum. But I am on the east coast but have seen about the same in GA TN AL over the last many years. ?ÿNGS Datasheet and OPUS are usually on same datum unless you have some older data sheets that still have ngvd29 on them. ?ÿOw I do realize that OPUS is Geoid derived NAVD88 and will not match exactly depending on Geoid quality in your area and such to the true benchmarks. 1 foot is a large separation. Is that a common occurrence in your area. ?ÿ

 
Posted : August 4, 2022 8:29 am
(@joe-the-surveyor)
Posts: 1948
Member
 

In my opinion, you are the professional signing the certificate.?ÿ If you are comfortable signing the cert based on the OPUS elevations, so be it.?ÿ In the end your signing the certificate puts you on the hot seat if something is not correct.?ÿ

In general I would say hold the NGS benchmarks, but I don't know where you are, the status of the bench, and all that goes with analyzing benchmark data.

The community official in my area is an architect, so I feel your pain when it comes to interpreting flood elevation certificates.

?ÿ

?ÿ

?ÿ

 
Posted : August 4, 2022 8:31 am
(@joe-the-surveyor)
Posts: 1948
Member
 

@olemanriver Could be a datum thing...in my area in CT it 1.08' between 29 & 88

 
Posted : August 4, 2022 8:36 am

(@lurker)
Posts: 935
Member
 

I'm just guessing the floodplain information was developed using NGS benchmarks. So if I wanted to provide relevant elevations to the floodplain it would make sense to compare apples to apples. If the floodplain information was based on OPUS observations then I would use OPUS. You have obviously shown that NGS and OPUS do not match. So using 2 different elevation realizations doesn't seem wise.

 
Posted : August 4, 2022 8:44 am
OleManRiver
(@olemanriver)
Posts: 2459
Member
 

@joe-the-surveyor That is my first hunch as well. I just did a certificate and the last time it was done and particular county still uses ngvd29. So about a foot. I checked to a few county monuments and a ngs benchmark. I was using network rtk. Ngs benchmark was .075 ft. And ngvd29 county monuments were 1.21 and 1.18. Using the ngs tools i did the transformation as a check and was whithen the conversion tolerances ?ÿof rtk and ngs nabd 88 to ngvd29. No statement of accuracy on the county mons. I placed navd 88 on certificate. Boss ls signed county said we need it on ngvd29 so he took the conversion and changed it. No big deal just have to have the metadata match the values etc. i use to know the rough difference from florida to Mississippi on that as i had to do support in those areas. Somewhere in the.5 ft to 1.6 ft depending on where someone was. Now that was a while back so my numbers might be wrong. Just using my memory. ?ÿ?ÿ

 
Posted : August 4, 2022 8:55 am
rover83
(@rover83)
Posts: 2346
Member
 
Posted by: @skeeter1996

I can see what elements are involved with OPUS derived elevations, but not NGS monument elevations.

The NGS is pretty good about putting the source of the orthos on the datasheet. It'll say how it was determined and when it was adjusted.

It's also pretty clear whether an ortho is NAVD88 or NGVD29, and if there are superseded value(s) they'll be on there as well. That can provide additional clues.

If an NAVD88 NGS benchmark (levelled and adjusted by the NGS per the datasheet) isn't matching OPUS-derived orthos using the latest geoid, then I would first be looking at whether there is a good published geodetic position (LLh) for that mark. If not, there's a good chance that the mark was not utilized in the latest hybrid geoid model and there could in fact be a discrepancy between what the geoid model says the separation is and what you are measuring ellipsoid vs the published ortho.

But one foot is a big difference. I don't know of any places in CONUS where the estimated uncertainty of the geoid is that large. Not that there couldn't be, I just haven't seen any as of yet.

So there could also be local movement/subsidence affecting the mark as well.

 
Posted : August 4, 2022 9:00 am
(@skeeter1996)
Posts: 1333
Member
Topic starter
 

@andy-j Craig, Montana. There is regularly a foot of difference in checked NGS monuments comparing between two monuments.

 
Posted : August 4, 2022 9:03 am
(@skeeter1996)
Posts: 1333
Member
Topic starter
 

@olemanriver There is about 3 foot difference in NGVD 29 and NAVD88.

I have found two NGS monuments to differ by over 4 feet between the monument versus their published Elevation. This has been the most extreme difference I have observed. In that instance the monument suspiciously is on a 8 inch pipe 4 feet high.

 
Posted : August 4, 2022 9:10 am

OleManRiver
(@olemanriver)
Posts: 2459
Member
 

@skeeter1996 where are you located. ?ÿGeographicallyƒ??. When you say between two monuments you see a difference not datum difference. ?ÿThat shows a disturbed monument of one or both assuming good measurement techniques were used. ?ÿ3 feet between ngvd29 and navd88. That is a difference in datums. Opus like above was stated is no where in conus using a geoid is that far off. So I guess whatever monuments you are holding are disturbed or subsidence etc. ?ÿNo matter which way you go having that information showing the difference in published vs opus you can easily show the county that your opus results are more realistic than the published marks value. And you can convert to ngvd29 from navd88 opus derived point. ?ÿIf i were licensed and i am not. I would have a nice 1 to 2 page little graphic and or report showing that the monuments on the ground do not check themselves and your opus results from two different sessions are better. ?ÿYou have proved movement on the ngs monuments if I understand correctly. So they are bad bad bad lol ????.?ÿ

 
Posted : August 4, 2022 9:22 am
Norman_Oklahoma
(@norman-oklahoma)
Posts: 7631
Member
 

Not all NGS datasheeted elevations are created equal. Are we talking about benchmarks with levelled elevations?

Also, are we talking about redundant 2+ hour OPUS-S solutions? ie/ have you eliminated the possibility of a measure up blunder? If so I'd have more confidence in the OPUS results.?ÿ

It is possible that the geoid modelling in the area is flawed.

 
Posted : August 4, 2022 9:31 am
(@skeeter1996)
Posts: 1333
Member
Topic starter
 

@rover83 What purpose would an Ortho have in establishing a NGS benchmark?

The benchmarks were established years before Orthos were developed. The benchmarks to my knowledge were established before any of that technology was developed. Most benchmark haven't haven't been located since they were set and none have had their ave had their elevations verified?ÿ

?ÿ

i

 
Posted : August 4, 2022 9:49 am
(@skeeter1996)
Posts: 1333
Member
Topic starter
 

@norman-oklahoma The geoid model has something to do with it I believe. That's another statistical use of a computer to come up with finite answers.

 
Posted : August 4, 2022 9:53 am
(@skeeter1996)
Posts: 1333
Member
Topic starter
 

@norman-oklahoma I have more trust in the OPUS Elevations. Redundant measurements have always matched within .003 meters. The error analysis done by OPUS is always in the .000 range. It's a fixed height tripod. Any measure up error would be less than .03 feet.

 
Posted : August 4, 2022 9:59 am

Norman_Oklahoma
(@norman-oklahoma)
Posts: 7631
Member
 
Posted by: @skeeter1996

The error analysis done by OPUS is always in the .000 range.

I get OPUS-S orthometric elevations with error estimates in the 0.06m range. Ellipsoid elevation errors are in the millimeter range. That's the geoid model.?ÿ

 
Posted : August 4, 2022 10:09 am
MightyMoe
(@mightymoe)
Posts: 9937
Supporter
 

you're stuck with the bench marks shown on the FIRM, if they show them as control.

I'm skeptical that the bench marks have much of anything to do with the data they used unless I was involved with it.

There is so much that goes into the BFE determination, you have to know the sequence of events.

How were the panels for the photos controlled if they did a community map, how were the sections collected for the valley cross-sections. It all counts.

I would really hesitate to dismiss a bench mark that's on the FIRM map. You say it's lower than CORS so it results in the structure being below BFE and if you use CORS you're above.

It's meaningless if CORS wasn't the basis of the contour data and you're taking on that liability. Was CORS even there when the mapping was done?

There wasn't a CORS point within 150miles when our local mapping was flown for FEMA control.?ÿ

 
Posted : August 4, 2022 10:12 am
rover83
(@rover83)
Posts: 2346
Member
 
Posted by: @skeeter1996

What purpose would an Ortho have in establishing a NGS benchmark?

Well, considering the NGS benchmarks reference orthometric elevations, I'd say a lot.

Posted by: @skeeter1996

The benchmarks were established years before Orthos were developed.

I don't understand this statement. Geodetic levelling campaigns were (and still are) used to establish ortho elevations for NGS benchmarks.

?ÿ

Posted by: @skeeter1996

The geoid model has something to do with it I believe. That's another statistical use of a computer to come up with finite answers.

How do you think they got final orthometric elevations from the original NAVD88 levelling campaigns, if not with a computer and utilizing statistics?

?ÿ

For what it's worth, GEOID18 computations place the 95% confidence error at +/- 4cm in the Craig area:

image
 
Posted : August 4, 2022 10:17 am
john-hamilton
(@john-hamilton)
Posts: 3349
Member
 

L 435 (SS1269) near Craig, MT. Cooperative Base Network Station/BM...occupied by OPUS twice in 2021...

Published NAVD88: 1054.011 m (leveled first order)

Published Ellip H: 1040.669 m (network accuracy 2.29 cm)

GEOID18: -13.342 m

Ellip H-GEOID18: 1054.011 m?ÿ

OPUS1 NAVD88: 1054.000 m

OPUS2: NAVD88: 1053.993 m

Seems like everything agrees very well. Disk in rock outcrop...

 
Posted : August 4, 2022 10:19 am
(@skeeter1996)
Posts: 1333
Member
Topic starter
 

@mightymoe There are no benchmarks referenced on the newer 2013 era FIRM maps or anywhere else in the FEMA documentation. The older NAVD29 maps did have benchmarks designated on them that were done by private contractors, not NGS.

I won't get into how FIRMS and BFE's were developed. That's another different subject.

My question is how do you relate the Elevations I find on the ground to the Elevations on a FEMA FIRM map most accurately.

 
Posted : August 4, 2022 10:22 am

Page 1 / 3