In the absence of an explicit standard for a job that requires obtaining horizontal and/or vertical control from known benchmarks in the vicinity, and where there is freedom on the part of the surveyor to choose how he establishes control on the project, what is the best/most expedient way to establish horizontal and vertical control? Any good how to manuals on the subject?
In my area, CORS stations are typically at 25 miles or more from a project. A single dual frequency GPS receiver set up for 2 to 4 hours, in this area, with a good view of satellite vehicles, will, by my experience, generate (by OPUS or other post processing software) solutions, constrained by CORS stations, in the vicinity of 2-3 cm horizontal and 6-10 cm vertical.
I'm wondering how that could be improved upon. I am aware that longer occupation times will theoretically improve the results, so that is one option to consider. However, will using additional dual frequency receivers in the vicinity of the project theoretically improve upon those results?
I understand that there is a limitation on the quality of the vertical solution based on the current geoid model in use which is still to be improved upon based on gravimetric modeling currently underway, so the vertical solution (i.e. orthometric height) will have its limitations. That is, no amount of additional observation time will eliminate the present errors associated with the geoid model.
What about the use of passive benchmarks that lie closer to the project than the CORS stations? In my area, there are passive control marks, typically 3rd order benchmarks (both horizontal NAD27 and vertical NGVD29) that were set in the middle part of last century, which have published coordinates to the hundredth of a foot, and which are often located within a few miles of a project. Some of these have since been destroyed, but some remain. These marks are often far enough away from the project that locating them directly by traditional leveling and/or traversing methods would prove time consuming and expensive, but it could theoretically be done. These passive benchmarks were typically set during a time period when GPS did not exist, and so often are not ideal locations for the setup of a GPS receiver. And, there remains the fact that the published coordinates for these marks also contain errors that I suspect will exceed the errors obtainable by GPS methods. Nonetheless, one method I can envision is to recover one or two of these benchmarks in the vicinity of the project, and to occupy with dual frequency receivers a couple of nearby points (in the vicinity of each benchmark) where there is a good view of the sky, and to use these GPS observations to constrain the project control by the passive benchmarks (local level and/or traverse loops would need to be run in order to tie the GPS observation points to the benchmarks). Does this have merit? What if there happens to be 2nd or even 1st order passive benchmarks closer to the project? For example, there is a 1st order vertical benchmark at a distance of about 10 miles from my typical project. It's on the wing wall of a bridge, and so won't be able to occupy directly with GPS receiver. Could I use the published elevation of this point to improve upon the solution at my project?
I know this is a rather rambling post, but I would appreciate any thoughts or comments. I am still learning about GPS, so please take it easy on me.
The older passive benchmarks usually have published values based on NGVD 29 datum. You would need to use a minimum of 2, 3 to get residuals to do an update to NAVD 88. We have performed several of these updates at dams throughout WV, OH, VA and KY.
So, would that be especially useful in cases where there was prior mapping for the project. As in, if the dam projects you mentioned were designed on the NGVD29 datum, using one or more of those same benchmarks, then the process you describe would enable holding that original datum?
Al,
You asked some good questions. Starting with your last sentence - it appears you have an exciting path ahead.
At the risk of providing more than you want or need, I'll offer the following comments with the understanding you have the option of all, part, or none of them.
1. Don't be afraid to ask questions - you've already demonstrated admirable traits.
2. Know your equipment and its limitations. You speak knowledgeable about features and capability.
3. Don't ever base your control survey on unproven or unreliable control points. Time saved by using second or third order (nearby) points is often wasted in the long run in time and frustration by trying to make the pieces fit.
4. Many do it, but I am skeptical of using OPUS in place of geodetic control. A static survey tied into known reliable HARN points (or CORS if available)can provide provable reliable points to be used for subsequent survey operations.
5. Elevations are difficult to achieve if done haphazardly. How much of your reputation are you willing to base on geoid modeling. It is well proven that "relative" geoid modeling values are better than "absolute" values.
6. I've posted several links to examples etc that may be helpful.
A. Example of static GPS survey at NMSU:
http://www.globalcogo.com/nmsunet1.pdf
B. Example of using a 3-D GPS survey to develop a 2-D plat (note, no map projection or "localization" is needed to obtain local tangent plane coordinates, directions, or distances - and standard deviations of same.):
http://www.globalcogo.com/3DGPS.pdf
C. Example of using GPS to establish a "first-order" elevation on a local HARN point.
http://www.globalcogo.com/ReilElev.pdf
D. Summary of procedures that have been used successfully to establish elevations using GPS. Note, relative geoid modeling values are used instead of absolute geoid modeling values.
http://www.globalcogo.com/gps-elev.pdf
If you have further questions, just ask - either in this forum or by direct email.
Earl F. Burkholder, PS, PE, F.ASCE
I've seen quite a few old disks (1934 etc) which have ADJUSTED NAVD88 elevations; I presume the NGS obtained these by running the original observations in a new adjustment based on NAVD88. The NGVD29 value is on the data sheet as superseded.
YMMV
I agree, be wary of OPUS.
I would suggest is to look at the CORS stations data in the area, mainly the velocities. NGS just published this data last fall. I'm down on the Cape and we have seen subsidence that has changed my NGVD - NAVD conversion by as much as 0.38'. Up in Vermont, they are still experiencing post-glacial iso-static rebound or uplift which will change your conversion in the other direction.
Unfortunately, we lost our one CORS station, so I have been given the opportunity to run my own static surveys. Just check and re-check and check into monuments and have some fun with it.
Earl,
Thank you for your post. A real gold mine. The links to the attached articles that you have written look to be excellent resources, and I will be spending some time reviewing those in detail in the coming days. That is a real help to me. Thank you.
I see that you have utilized HARN points in your solutions. Where can I find a list of HARN points in my area?
Al
I'm not crazy about control. Don't get me wrong, I do it right, but in NC we have standards that are published. As long as I am in accordance with the standards I'm good! GPS is great for 99% of the control I do (horizontal and vertical). If you need tighter vertical control, find a 1st order monument and use a level. Good Luck!
It all depends on project requirements. When I first started doing GPS, we only had NAD27/NGVD29. Even so, I would only use NGS published mons, not USGS. Why? Because the USGS control is only 1:10,000 H and third order V. So if a USGS triangulation station was 10 miles away from an NGS chain, it could be 5' off and still be within specs. Some were further away than that. We had a lot of eccentrics to triangulation stations in the woods here in the east.
If a point does not have NAD83/NAVD88, then NGS didn't use it/adjust it. I would avoid those at all costs. Typically, USGS or USACE.
UNLESS>>>>>
Many projects I work on have their own control networks established in the 30's or 40's, referenced to NAD27/NGVD29, often USGS mons. So, you can't just do a NADCON or VERTCON transformation and expect it to agree very well. We recover as many points as we can, survey them in with GPS, and compute a transformation between the old system and the CORS/HARN.
If there are no requirements, then I bring H in from CORS (and V if not required to be super accurate). If good NAVD88 is needed, I look for 3 or 4 stable benchmarks surrounding the project, and static GPS them in.
A recent project was to reset some boundary monuments. These had NAD27 coordinates, and each point had references to nearby trees. But, the source of the NAD27 was not known. The best thing to do was to find adjacent mons, occupy them, and traverse over to reset the missing mons, checking the references at each one. It would not have worked to try and use GPS with NADCON, as that would not be close enough.
So, unless there is a requirement to absolutely use NAD27, I would never use them, as there is too much "slop" in the old work.
NADCON/VERTCON is never close enough to use for anything other than recon in a situation like you describe. We are actually working on several of those boundary remonumentation surveys now that have mapping based on NAD 27. CORPSCON those to NAD 83 for recon but any that need reset require a process like you described.
> In the absence of an explicit standard for a job that requires obtaining horizontal and/or vertical control from known benchmarks in the vicinity, and where there is freedom on the part of the surveyor to choose how he establishes control on the project, what is the best/most expedient way to establish horizontal and vertical control? Any good how to manuals on the subject?
>
> In my area, CORS stations are typically at 25 miles or more from a project. A single dual frequency GPS receiver set up for 2 to 4 hours, in this area, with a good view of satellite vehicles, will, by my experience, generate (by OPUS or other post processing software) solutions, constrained by CORS stations, in the vicinity of 2-3 cm horizontal and 6-10 cm vertical.
>
> I'm wondering how that could be improved upon. I am aware that longer occupation times will theoretically improve the results, so that is one option to consider. However, will using additional dual frequency receivers in the vicinity of the project theoretically improve upon those results?
>
> I understand that there is a limitation on the quality of the vertical solution based on the current geoid model in use which is still to be improved upon based on gravimetric modeling currently underway, so the vertical solution (i.e. orthometric height) will have its limitations. That is, no amount of additional observation time will eliminate the present errors associated with the geoid model.
>
> What about the use of passive benchmarks that lie closer to the project than the CORS stations? In my area, there are passive control marks, typically 3rd order benchmarks (both horizontal NAD27 and vertical NGVD29) that were set in the middle part of last century, which have published coordinates to the hundredth of a foot, and which are often located within a few miles of a project. Some of these have since been destroyed, but some remain. These marks are often far enough away from the project that locating them directly by traditional leveling and/or traversing methods would prove time consuming and expensive, but it could theoretically be done. These passive benchmarks were typically set during a time period when GPS did not exist, and so often are not ideal locations for the setup of a GPS receiver. And, there remains the fact that the published coordinates for these marks also contain errors that I suspect will exceed the errors obtainable by GPS methods. Nonetheless, one method I can envision is to recover one or two of these benchmarks in the vicinity of the project, and to occupy with dual frequency receivers a couple of nearby points (in the vicinity of each benchmark) where there is a good view of the sky, and to use these GPS observations to constrain the project control by the passive benchmarks (local level and/or traverse loops would need to be run in order to tie the GPS observation points to the benchmarks). Does this have merit? What if there happens to be 2nd or even 1st order passive benchmarks closer to the project? For example, there is a 1st order vertical benchmark at a distance of about 10 miles from my typical project. It's on the wing wall of a bridge, and so won't be able to occupy directly with GPS receiver. Could I use the published elevation of this point to improve upon the solution at my project?
>
> I know this is a rather rambling post, but I would appreciate any thoughts or comments. I am still learning about GPS, so please take it easy on me."""
you might also want to consider taking a level to your local passive marks, set an eccentric station that is GNSS 'friendly', then get a d.i.e. from passive mark to new (eccentric) mark, THEN start your gps session from eccentric station to your project.
my apologies to mr burkholder, if necessary, as i have not the time to read his links. he's probably addressed this situation in one of the sources quoted.
great day to all, Eddie Glawe
Al,
It appears you are getting lots of good help from other postings as well.
You are probably aware of the NGS web site for lots of good information on GPS surveying:
http://www.ngs.noaa.gov NGS home page
From the drop-down menus under "data & imagery" you can find
1. Data sheets for individual stations.
2. GPS data (CORS)
A link for State Advisors is given on the left side of the NGS home page. Dan Martin in Vermont is probably the closest to you. I wouldn't avoid contacting him just because I live in a different state.
But probably the neatest feature for locating and displaying NGS marks is the DS World progam that works in conjunction with Google Earth. I believe you can find DS World at http://www.ngs.noaa.gov/PC_PROD/PARTNERS/index.shtml. (I just checked and the page was not found - sorry about that. On the NGS home page do a search on "DS World" and follow the link to user-contributed software for DS World)
Good luck and happy exploring.
Earl F. Burkholder, PS, PE, F.ASCE
Passive Network (HARN)
The NAD83 Coordinate estimates on the Passive Network Stations (HARN, HPGN, FBN, CBN etc.) will change in a couple of months (about April of this year)!
See:
http://www.ngs.noaa.gov/web/surveys/NA2011/
My experience with the Passive Network has on average been very good (a few exceptions), but you must bear in mind that the Passive Network is a “snapshot in time” that may or may not be valid today. Most of the “HARNS” around my area were last observed in 1999, and some were last observed in 1994 (or even earlier).
The CORS (Active Network) is observed 24/7/365! Using the 60 Day and Long Term Time Series, you can get a pretty good idea of how stable a given CORS is, and “accurate” the current Coordinate Estimate is relative to Datum.
Just my 2-bits.
Loyal
d.i.e. is "difference in elevation"?
Passive Network (HARN)
I agree with Loyal on this, Passive BMs are contingent on stability. Many of these Marks haven't been revisited in anywhere from 20-80 years. I believe the Navd88 els were computed based on the previously published data.
Ralph
Passive Network (HARN)
Right on Loyal. Another good thing to check with your prospective CORS station is if it has computed vs. modeled coordinates. You can find the list of the computed CORS here. The computed stations have a minimum of 2.5 years worth of data.
Doug
yes sir
Passive Network (HARN)
While my comments above related to the “HARN” Horizontal [GPS] part of the Passive Network (some are also Bench Marks of course), Ralph's comments about Bench Marks per se is RIGHT ON.
Even “Stability A” Bench Marks can (and do) move relative to Datum, and when you start talking about less stable settings (e.g. 'C' & 'D') then all bets are off. EVEN IF the circuit is relatively “fresh” (which around here would be AT LEAST 28 years, and usually more like 50 or 60 years), you can EXPECT a certain amount of differential movement within a given circuit.
Fitting EXISTING Control is one thing, establishing NEW (and repeatable) Control for the future, is another.
Loyal
VERTCON
In the documentation, the solution that VERTCON provides is not appropriate for 1st, 2nd, or 3rd order elevations. 3rd order is equivalent to 0.05' per root mile (square root of miles) which is the minimum I've seen in State Requirements. It is great for mapping and GIS, but not for our work...
By the way, for those folks getting new FIRM maps this year, be vigilant! The FIS I reviewed states that the conversion will be 0.9', no ifs and or buts, then in small print it directs you to the NGS for a conversion. In my neck of the woods, the VERTCON conversion is around 0.9' but the actual conversion is around 1.1'.
Now, can I stake out a contour where the FIRM tells me 10.0 and I actually stake out 9.8? That would flip the Conservation Commission out! They already were upset with my plan on NAVD where the mean high water elevation was off by 1.1'.
Ignore the Sig Figs, the FIS figure is listed to one place and I just through out my conversion to one place.
Just spent some time looking for DS world and ran into a 1.5 hr video on RTK produced by NGS, hosted by Bill Henning. Watched the whole thing, with great interest. Fantastic resource.
http://www.ngs.noaa.gov/corbin/class_description/RT_field_2011.shtml
