Kent McMillan, post: 393379, member: 3 wrote: the relative elevations of pipe inverts at adjacent nodes in the network is quite a bit more important than those several nodes removed
Assuming this to be correct, the leveling option looks like the preferred technical approach.
At the client's request, I ran a couple of 0.2' scenarios to see if it would bring the cost down significantly. One relied on trig leveling (total station), the other on RTK. Both came out a bit more expensive than 3rd Order leveling.
RTK might have won the cost battle in this case but for a couple of things:
1. This is a vertical-only project. SPC positions for the manholes already exist.
2. Much of the sewer system layout lends itself to level loops, so leveling blunder detection is available without double-running for those portions. On the other hand, I'd need at least 2 time-separated RTK observations to consider a vertical number reliable.
Jim Frame, post: 393400, member: 10 wrote: Assuming this to be correct, the leveling option looks like the preferred technical approach.
Yes, leveling with a digital level sounds like a winner to me on that one. I assume that you level to two opposite points on the frame and then return later to pop the lid and measure to the invert?
Kent McMillan, post: 393403, member: 3 wrote: I assume that you level to two opposite points on the frame and then return later to pop the lid and measure to the invert?
For general topo I shoot the center of lid unless it's noticeably sloped. For this project -- if I get it -- I'll probably punch the rims and reference the inverts I do measure to that. (Of the 620 or so manholes involved, rim-to-invert distances exist for all but 89 of them.)
Jim Frame, post: 393416, member: 10 wrote: (Of the 620 or so manholes involved, rim-to-invert distances exist for all but 89 of them.)
That does sound like a pretty "clean" job with digital level and Star*Lev.
Kent McMillan, post: 393418, member: 3 wrote: That does sound like a pretty "clean" job with digital level and Star*Lev.
Maybe!
Unless things have changed in recent years, Star*Lev does NOT perform orthometric corrections of any kind. This may or may not be a big issue on smaller projects, but on something 7 miles x 3.5 miles in California, this MAY be significant (especially considering the tolerances requested).
Just say'n
Loyal
You can get a pretty good handle on the 'what-if' by using the NGS program LVL-DH (linked below) and the 3 Bench Marks mentioned in the opening post.
http://www.ngs.noaa.gov/cgi-bin/lvl_dh.prl
You will also want to check a mile or two each side because that may be the "kicker" if you are running up/down a valley. It's an easy check, and it's always better to be safe than sorry.
Loyal
The gravity field is pretty consistent around here, so I don't think the ortho height correction is going to matter. I'll give it another look, though.
Jim Frame, post: 393492, member: 10 wrote: The gravity field is pretty consistent around here, so I don't think the ortho height correction is going to matter. I'll give it another look, though.
Jim,
Do you have the NGS PID(s) on those 3 Bench Marks?
Loyal
AI5051
DK4487
JS4556
Jim Frame, post: 393500, member: 10 wrote: AI5051
DK4487
JS4556
You're right Jim, the ortho corrections are ZERO between these stations.
On the other hand, ALL three stations are GPS derived NAVD88 heights rounded to the nearest centimeter! I would HOLD ONLY one of them, and just hope to hit the others within a centimeter (or so). Just my opinion, but I would NOT "balance" a digital level run between them, but would be comfortable HOLDING any ONE of the three (in the larger sense). There are REAL Bench Marks in the area (according to the NGS IDB) which I would try and tie into, but that's just me.
Loyal
Loyal, post: 393502, member: 228 wrote: ALL three stations are GPS derived NAVD88 heights rounded to the nearest centimeter!
I know, I headed up the project that established their bluebooked heights. (I had a lot of help from Don D'Onofrio and Mike Potterfield.)
Loyal, post: 393502, member: 228 wrote: I would NOT "balance" a digital level run between them, but would be comfortable HOLDING any ONE of the three (in the larger sense).
Assuming they hold together somewhere around their intended error limit (2 cm), I'll probably pick something in the middle but adjust the leveling only on itself.
Loyal, post: 393502, member: 228 wrote: There are REAL Bench Marks in the area (according to the NGS IDB) which I would try and tie into, but that's just me.
This is in a subsiding area, so the elevations of leveled stations in the area aren't reliable. We reached out about 40 km to the east to get a bedrock NAVD88 First Order mark. There are no NAV88 leveled marks to the west on this side of the Vaca Mountains, so we Vercon'd a stable NGVD29 leveled mark at the base of the hills about 30 km west of this manhole project as a check. The rest is pure geoid model.
Jim Frame, post: 393505, member: 10 wrote: I know, I headed up the project that established their bluebooked heights. (I had a lot of help from Don D'Onofrio and Mike Potterfield.)
Assuming they hold together somewhere around their intended error limit (2 cm), I'll probably pick something in the middle but adjust the leveling only on itself.
This is in a subsiding area, so the elevations of leveled stations in the area aren't reliable. We reached out about 40 km to the east to get a bedrock NAVD88 First Order mark. There are no NAV88 leveled marks to the west on this side of the Vaca Mountains, so we Vercon'd a stable NGVD29 leveled mark at the base of the hills about 30 km west of this manhole project as a check. The rest is pure geoid model.
I hear ya Jim.
It is a problem in most (if not all) of the valleys (basins) West of the Rockies (and many areas East of them too). Even the "Bedrock" Bench Marks are moving around to some extent of another. Coupled with the "stale" leveling in most of the West, it can be a real crap shoot.
Speaking of Mike P., have you talked with him lately? It's been nearly a month since I have, and he seems to be enjoying retirement.
Loyal
Loyal, post: 393509, member: 228 wrote: Speaking of Mike P., have you talked with him lately?
I traded email with him a week or so ago. I had heard he was going to adjust the reobservation of a project we partnered on in 2004, and wanted to touch bases. It turns out he's not involved, and doesn't plan to be. I guess retirement is treating him well.
Loyal, post: 393487, member: 228 wrote: orthometric corrections ... on something 7 miles x 3.5 miles in California, this MAY be significant (especially considering the tolerances requested)
I guess it doesn't matter on this project, but for critical tolerance on fluid flow, I would think they should work in dynamic heights instead of orthometric heights. That requires a sort of opposite correction formula. The old thing about water (and whatever) running "uphill" according to orthometric numbers, whereas dynamic tells the true story.
Bill93, post: 393517, member: 87 wrote: I guess it doesn't matter on this project, but for critical tolerance on fluid flow, I would think they should work in dynamic heights instead of orthometric heights. That requires a sort of opposite correction formula. The old thing about water (and whatever) running "uphill" according to orthometric numbers, whereas dynamic tells the true story.
Agree!
In fact, IF one really wants the best results, you would use the BEST gravity model available (not the NAVD88 gravity Model), and compute geopotential numbers for the various TBMS and project points. From there the dynamic heights are a breeze (although a little subjective as far as units go).
Loyal
Bill93, post: 393517, member: 87 wrote: I guess it doesn't matter on this project, but for critical tolerance on fluid flow, I would think they should work in dynamic heights instead of orthometric heights. That requires a sort of opposite correction formula. The old thing about water (and whatever) running "uphill" according to orthometric numbers, whereas dynamic tells the true story.
I'm trying to think of a public works project involving fluid flow over a distance of about 7 miles where the difference between dynamic and orthometric heights would be significant. As I recall, in open channels the unknowns in the roughness or friction coefficients used in the models are much, much larger than any change in the slope due to dynamic-vs-orthometric height differences could possibly cause.
Kent McMillan, post: 393523, member: 3 wrote: I'm trying to think of a public works project involving fluid flow over a distance of about 7 miles where the difference between dynamic and orthometric heights would be significant. As I recall, in open channels the unknowns in the roughness or friction coefficients used in the models are much, much larger than any change in the slope due to dynamic-vs-orthometric height differences could possibly cause.
Kent,
While I agree with your comments above, when a Professional Surveyor "Certifies" to a particular "accuracy/precision" without fully contemplating all of the underlying nuances between NGVD29, NAVD88, Ortho, dynamic, etc. etc., it's generally not a good thing.
Jim obviously knows what (and where) he is going with all of this, whereas the (not so) subtle differences in "heights" become non-trivial very fast in some parts of the country.
Practical matters (reality) aside, the "spec" didn't ask for an engineering opinion, it asked for a ridiculous (IMO) suite of "elevations" spread over a rather large area.
Just say'n
Loyal
Loyal, post: 393528, member: 228 wrote: Kent,
While I agree with your comments above, when a Professional Surveyor "Certifies" to a particular "accuracy/precision" without fully contemplating all of the underlying nuances between NGVD29, NAVD88, Ortho, dynamic, etc. etc., it's generally not a good thing.
Just out of curiosity, what do you think the maximum error introduced by ignoring the orthometric height correction would be in a 3.5-mile line of leveling between two benchmarks in a typical urban setting such as that of Jim's project? The premise is that the line is adjusted by holding the elevations of both as published in the NGS database as conditions of the adjustment and also assuming no mark movement is involved, i.e. that the published elevations do represent the elevations of the marks.
Kent McMillan, post: 393529, member: 3 wrote: Just out of curiosity, what do you think the maximum error introduced by ignoring the orthometric height correction would be in a 3.5-mile line of leveling between two benchmarks in a typical urban setting such as that of Jim's project? The premise is that the line is adjusted by holding the elevations of both as published in the NGS database as conditions of the adjustment and also assuming no mark movement is involved, i.e. that the published elevations do represent the elevations of the marks.
I don't know Kent...
It depends on what you mean by "urban!" Park City Utah is "urban," but has over 600 ft. vertical along the roughly 2 miles of "Main Street."
ASSUMING that two "Published" Bench Marks (regardless of the distance or height difference between them) have no post-adjustment variance in my neck of the woods is insane at best. Unless you apply the proper corrections to your observed data, you have no way of "knowing" (within reason), whether the tenth of a foot (or more) of variance between your "leveled difference" and the Published Difference, is real or not. That is why I suggested running LVL-DH between the BMs. More importantly (sometimes) it's what happens BETWEEN the Bench Marks that will bite you.
Jim doesn't have to go very far East to bump into the gravitational influence of the Sierra Nevada Range, and at that point of the conversation, I didn't know exactly where his project was located.
I have never worked in pancake land, so I tend to see the world from the perspective of high elevations, mountains, and the Basin and Range Province. Aside from a few weeks in Northern Maine, and a week or so in Southeast Alaska, I have never really done any work at low elevations. Even my work in California, Oregon, & Washington was in the Cascade or Sierra Nevada Mountains.
Loyal
