Both of those issues are due to control and metadata screwups, not any inherent problem with geoids or ellipsoid heights.
Both of those issues are due to control and metadata screwups, not any inherent problem with geoids or ellipsoid heights.
Annnnd the proposal is to continue using ellipsoid and geoid models instead of base NAVD88 elevations. Which for many tasks I refuse to do. I'm much more precise that that.?ÿ
?ÿ
We had two rather spectacular FUBAR's from people using OPUS/Geoid model derived elevations.
Neither of those two situations can be directly attributed to using OPUS derived orthometric elevations.?ÿ They can be attributed to using elevations derived form a source independent of the original project basis and left unconfirmed. I've encountered numerous times when one brass benchmark disagreed with another.?ÿ
Plenty of times I've done redundant OPUS sessions on the same monument and come up with elevations in a 0.2' range. The geoid separations employed would be identical for each session, the ellipsoid elevation varied. That is the limit of precision of the method.?ÿ
In the 2nd case in particular - who does such a pre-design topo survey and doesn't confirm the finished floor elevation of the building to be attached to??ÿ?ÿJust very bad practice.?ÿ
Annnnd the proposal is to continue using ellipsoid and geoid models instead of base NAVD88 elevations. Which for many tasks I refuse to do. I'm much more precise that that.
Yeah, about that precision...from the NGS itself:
GEOID18 has an estimated uncertainty of about 1.5cm relative to NAVD88.
?ÿ
?ÿ
As with some others as described above, my approach has always been to maintain the observation data and adjustment criteria.?ÿ?ÿ
Thanks for posting Jim; I thought I was the only one. I don't database my processed or adjusted data like many have posted about above, only my raw data. Each job has a different user, criteria, and purposes. I routinely reprocess older observation data from previous nearby jobs to keep from redoing the same. You said it better; Yea, what you said.
?ÿ
Annnnd the proposal is to continue using ellipsoid and geoid models instead of base NAVD88 elevations. Which for many tasks I refuse to do. I'm much more precise that that.
Yeah, about that precision...from the NGS itself:
GEOID18 has an estimated uncertainty of about 1.5cm relative to NAVD88.
?ÿ
?ÿ
That's really good. It's not quite there for us locally, but it's waaayy better than it was when 10cm was the norm. If I had it working 1cm or less I'd not worry at all about it. I'm very happy with results we are getting which is 2-3cm. That's great, but not what you're seeing either from that paper or where you work.?ÿ
?ÿ
Yeah, about that precision...from the NGS itself:
There are exactly two NGS levelled benchmarks in my city of 20 square miles, and no others within a mile of any border thereof.?ÿ These benchmarks are about 2 miles apart "as the crow flies" and were last levelled in the 1930's.?ÿ ?ÿ One is in the doorsill of an old building, the other in the concrete wingwall of a culvert passing under a railway. They look OK, but who knows? Oh, there is about 900 feet of relief within the city limits.
The county has a benchmark network but nobody really knows anything about how or when it got there. It was old as long as anyone can remember. The original list of descriptions and elevations is typewritten and well thumbed.?ÿ The monuments are of various quality, some in curbs upthrust by tree roots, others in sunken sidewalks. And some look fine. But who knows? Today, about a dozen monuments exist within the city limits. The elevations are given only in NGVD29.
There are some state highways running through our fair city and some of the overpasses still have some really impressive looking brass disks in them. Once upon a time you could get a location description and record elevation from the DOT, but no more. The data for them is no longer available, except in cases where these same disks were also made a part of the county's network. The DOT has abandoned their use. They bring elevations to their projects through the use of the VRN they operate, and level between control points.?ÿ ?ÿ
My city has a "network" of PK nails in curbs and a typewritten list of elevations that covers about 1/4 of the area - that is, the part that was within city limits around about 1970. No metadata at all.?ÿ ?ÿ
The story is similar in all the adjacent cities and counties, as it is everywhere I've worked. One city in Canada, for example, was so bad as to be comical.?ÿ They had established their benchmarks using summer interns with, apparently, virtually no supervision - discrepancies of several feet were the rule.?ÿ?ÿ
In short, there are no passive marks that you can point to and be absolutely certain that they are correct. Youi find one that looks good and you use it. You confirm it as best you can. You be sure to state which one you used on the face of your topo map. And if you are following up someone else's work, you use the same one and check, check, check.
Under these circumstances I'm comfortable using OPUS and Geoid models. At least it is being actively managed.?ÿ ?ÿ ?ÿ
Yeah, about that precision...from the NGS itself:
...snip
The story is similar in all the adjacent cities and counties, as it is everywhere I've worked. One city in Canada, for example, was so bad as to be comical.?ÿ They had established their benchmarks using summer interns with, apparently, virtually no supervision - discrepancies of several feet were the rule.?ÿ?ÿ
?ÿ...snip
?ÿ
If you do not want to publicly shame them, I kinda get it.......
?ÿ
?ÿ
?ÿ
..........but could you PM me the city so that I know to never base any of my vertical checks off their network? 😉
?ÿ
..........but could you PM me the city so that I know to never base any of my vertical checks off their network?
Chilliwack, B.C. 1994.?ÿ It was no secret. I'd suppose they have done something about it by now. But maybe not!?ÿ
It's interesting to hear about lack of bench marks in more urban areas. I've worked in a number of nearby towns and all of them have some form of elevation control, some better than others.
We even helped a colleague establish control for a small town some 300 miles away since the town wanted to establish a base for on-going construction projects.
Even that town has a set of first order bench marks running along the railroad that skirts the town limits. Those proved to be very useful combined with the GPS survey he did.
This was early on in GPS days pre Geoid03 so the GPS derived elevations were awful. However, it did lead us to find a bad bench mark that must have been disturbed during some construction in the past. Running levels through it showed it to be .6' out.?ÿ
Locally there are still more than a dozen first order bench marks, only one that has been identified as disturbed. There are plenty of city bench marks first expanded during an Army Corps project, then a water treatment plant, then an extensive interceptor sewer project. The city has a bench mark booklet, a set of plans for as built elevations for the sewer line and many other manholes and valves. The many street improvement projects are also available, bridge reconstructions, airport surveys with bench marks at the end of the runways, two first order bench marks at the airport. Each and everyone I've ever run levels through check very well.?ÿ
We just completed a survey for a Condo project and recovered 5 manholes, they all checked less than .1' to the given elevations for the rims on the city plans.?ÿ
There is a set of pre NGVD29 marks scattered about. They are large, usually extend at least 1-2' out of the ground and are stamped with elevations that are about 10 feet different than 29 elevations. Bench Marks and elevation control has never been much of an issue locally or for nearby municipalities.
Just to be clear, these are the datasheet numbers for the first order Bench HARN point:
airport surveys with bench marks at the end of the runways, two first order bench marks at the airport
Twenty years ago I used to be able to drive out to the county airport and tie into their control at will, but now the whole thing is fenced and getting permission to access any control there is a multi-day project in itself, involving multiple emails, phone calls and meetings.?ÿ It's not worth the trouble.
airport surveys with bench marks at the end of the runways, two first order bench marks at the airport
Twenty years ago I used to be able to drive out to the county airport and tie into their control at will, but now the whole thing is fenced and getting permission to access any control there is a multi-day project in itself, involving multiple emails, phone calls and meetings.?ÿ It's not worth the trouble.
Same here, fortunately, during one of the many construction projects there, a building was being expanded over a NGS point (also a Bench Mark) so it was moved out of the security area to an accessible public location. Makes it really nice to have that point. Unfortunately, a bench mark on the side of the terminal building was relocated to a corner of the terminal on the security side of the fence. Win one, lose one.?ÿ
?ÿ
I'm really hoping that public sector organizations will take a hard look at the costs of trying to stick with decades-old datums on passive marks as opposed to switching to the modernized NSRS for the future.
It's more work and more $$$ for us if they decide to stick with legacy datums and we have to go backwards and sort out discrepancies for every project...but I would just as soon everyone get on the same page.
I'm really hoping that public sector organizations will take a hard look at the costs of trying to stick with decades-old datums on passive marks as opposed to switching to the modernized NSRS for the future.
It's more work and more $$$ for us if they decide to stick with legacy datums and we have to go backwards and sort out discrepancies for every project...but I would just as soon everyone get on the same page.
I'm with you, once FEMA works through their mapping with 2022 it will bring everyone on-board. If it's like 29 to 88 FEMA will be the last step. I'm guessing DOT and Federal money projects may force it before FEMA, but who knows. The latest iteration of Geoid models work so well capturing the Geoid contours that it's really not a problem to use them exclusively. As long as you can locate two points a mile apart and level between them and get the same difference as GPS you'll be golden.?ÿ
?ÿ
this has been a great discussion.
I'm glad we're having it.
?ÿ
this site rocks!
