Hello all.?ÿ I am new to this site, so forgive me if this has been addressed somewhere and I just didn't find it.?ÿ I did stay up till 4 AM last night trying to find an answer and didn't find one.?ÿ So, I am a licensed engineer and a licensed surveyor.?ÿ I've performed field surveying for quite a few years over my career, but most of my career has been in design.?ÿ I tell you this so you understand that I'm not a novice, but I recognize that I don't have decades of true surveying experience like many of you probably do.?ÿ That said, I do know my way around a total station and a data collector and know how to do direct / reverse measurements and run a traverse correctly, etc.
So, here's my problem.?ÿ I am currently working on a survey project that requires me to establish control for a small (.25 acre) site where a municipality is going to build a roundabout.?ÿ I found two recorded points fairly close by and decided to traverse from each one to my site and establish a control point.
Point one is a NGS monumented disc with two reference marks adjusted to NAD83.?ÿ Point 2 is a State of Connecticut DOT disc with three reference marks in NAD27.?ÿ I was able to to locate the two reference marks for the NGS point but not the actual point itself.?ÿ For the DOT disc, I found it and all three reference marks with no problem.?ÿ Also, it should be noted I didn't use the DOT disc at first because I hadn't discovered it yet.
So, I started by using the NGS point.?ÿ I did a resection based on the two reference marks and establish a point.?ÿ It closed 1:42000 or thereabouts, so I believe that is fine for my particular application.?ÿ I then set a traverse point, turn to it, back sight, etc. and then from that point set my first control point, then set a second control point.?ÿ Everything checks in when I close the traverse so I pack up and go home.?ÿ The next day I go out and I set up on my second control point, back sight the first control point I set the previous day, and shoot points for the next two days, thereby completing the project, or so I believe.?ÿ I then head into the office and get everything all done and decide to overlay an orthophoto just to see how close I am and I about lose my lunch.?ÿ My survey was shifted from the ortho by about 5'.
So, I do some digging and I get the DOT disc and the I go out again to check my control using this point.?ÿ Now its off by like 12' in the other direction.
Here is some other relevant data:
1.?ÿ The instrument was calibrated in November.?ÿ So, I don't believe it is the instrument.
2. The length of the first traverse (NGS) was about 1200' with two setups with about 140' elevation change between the start and end.
3. The length of the second traverse (DOT) was about 800 with two setups with only about 30' elevation change between the start and end.
4. Both the NGS and the DOT points are third order horizontal accuracy points.
Why, oh why, are my points a) coming up so different relative to each other??ÿ My main control point I set from my first traverse is about 27' different than where it came up when I traversed from the DOT disc... WHY??? and b) so different from the orthophoto??ÿ What am I doing wrong?
Azimuth. How did you determine azimuth from the NGS point?
Azimuth again: grid v geodetic. Did you use a geodetic azimuth and either forgot to apply convergence, or applied convergence the wrong way? (eg: Add instead of subtract)?ÿ
I didn??t see any reference to how you advanced from the NGS (reset) point.
(The RMs for the NGS (reset) are not an accurate source of azimuth.)
i've had similar issues a few times in the past and it has invariably been datum issues.?ÿ as in, circle back and make sure your 27-83 conversion is correct, and go through the datum of every last source and make sure it's EXACTLY the same apples as the other apples.
also, don't know about your parts, but the DOT here has a funny habit of slipping projects into weird, bastardized datums for some reason or other.?ÿ?ÿ
iow- my immediate reaction would be to go back and read all the fine print i can find regarding bearing and coord bases.?ÿ and check the 27-83 thing.
Check you are using U.S. Survey feet instead of International Feet? This can easily cause an unexplainable 5-12' shift in coordinates.
Interesting. I just assumed the azimuth's given are geodetic. Certainly worth checking.
That is what I suspect too, but I feel like I've checked and rechecked. Not sure where else to find any fine print. If I can find out how, I'll attach the control cards I used. One note about the DOT point was that the reference mark azimuth's were weird. It too me a bit of time to get them lined up correctly. There was an error with the azimuth between RM 2 and RM 3 being reversed, ie, the azimuth for RM 2 that was recorded was really for RM 3 and vice versa.
Also check you converted the NGS point coordinates from Meters to US survey feet. Many calculators and converters are default to convert to INT. Feet.
Yeah, I did ensure that. In this particular area, that would account for about a 3' shift, but that is not the issue here.
And, bam, just like that, it was the convergance. Thank you! Lines up like a freaking perfect puzzle piece now.
What is the NGS point that you are using??ÿ
Conversion from NAD27 to/from NAD83 is always approximate and the accuracy varies with location.?ÿ The difference often being a meter or more.?ÿ You may be dealing with three different datum shifts as well as NAD83 shifting (Epochs) over time.
-
- NAD27 to NAD83 (Epoch ??) on DOT monument.
- Orthophoto NAD83 (Epoch ??)
- NGS (reset) monument (Epoch ??)
You may need to find out what physical control points were used for the aerial photo and tie your work to those. The NAD83 datum has been updated and although the shifts are small they may be a part of what you are dealing with.
And a pair of points, not far apart, and one nad27 converted, is a very very poor source of azimuth.
Usually, I hear about much larger differences than this.
Sometimes, the problem is with YOUR data, and sometimes the problem is with the PREVIOUSLY PROVIDED data.
Discrepancies are expected. The amount of discrepancy is a variable.
Maybe a predecessor used rtk, with "here" position for their base. Typically, this will be off 3 to 10 feet. If their bearings were geodetic, this could explain the rotation differences.
Stay with this, until you:
Know your data is good.
Where the discrepancy is.
How to properly relate the 2 data piles.
Can publish your meta-data, with knowledge of its meaning.
And, finally, can direct others on the correct path.
This is all part of what it means to "be a pls".
We will try to help. But, it may take a number of steps.
One more thing. IF one of those markers is disturbed, it could be handy to get some GPS out there.
Nate
?ÿ
I think even if you did everything correctly things probably wouldn't appear to line up.?ÿ Is the client making you use both of these monuments??ÿ Maybe the simple solution is to just ditch one of them and carry on with the project.
I would be highly suspicious of any conversion from NAD27 to NAD83 if you are looking for tight agreement with other data.?ÿ But if the DOT is designing with the coordinates on their point, you have to make sure you know exactly how they treated it and probably rely on it over the NGS point (which maybe better in the overall world).?ÿ If they gave you SPC coordinates to convert, either make sure they are true SPC and not rescaled to ground or else get the scale factor they used.
Is the NGS point in NAD83(86) or has it been updated with GPS data to a later realization?
I cannot see if you did enough field traverse work to get started in the direction you want to go. You have 2 NGS reference points, 1 DOT control?ÿ and 3 DOT reference points. All 6 must be included in your traverse to first verify the quality of your field work.?ÿ
Second I have never seen orthophotos more that 0.5' off my field work, of course my fieldwork includes GPS observations which is where I suggest you go next. If not take the field coordinates you have that are closest to orthophoto info and work on finding why there is a difference. I bring my orthophotos into CAD drawings and care is needed to get the photo referenced and properly scaled. Typically I get orthophotos in a zipped folder with metadata files included and sometimes even directions. I prefer tiff format but am happy with Mr. Sid if that is available.
To get better answers here, share more info on where you are, what you have (equipment, software and info) and how you plan on getting to the destination.
Paul in PA
This is a classic apples and oranges problem. I'm surprised you are only 12 feet off.
you have a DOT control point in some version of NAD27. Conversion from NAD27 to 83 is fraught with issues and doesn't necessarily match the NGS NAD83 you have.
You have an NGS control station in some flavor of NAD83. NAD 83 has datum tag and epoch which affects the resultant coordinates.
Then you have an orthophoto in some version of an unstated datum which may not match the other two.
You need to know what the orthophoto is controlled by to match it.
I don't use any NGS points unless they have been occupied by NGS with GPS, which the data sheet will explain. The phrase adjusted to NAD 83 makes me suspect that it's an old NAD27 point that has never been re-occupied.?ÿ
The NAD83 value then is a conversion and Corpscon will warn you that the conversion will be "off".
As stated above the NAD 27 value for the DOT point is probably useless for tight control work unless you use it in conjuction with other NAD27 control points in the same project,,,,,,,and stay in NAD27.
A nice monument that needs me to establish good coordinates for it.
First thing I would try would be to take the NAD 27 coordinate for the NGS point, the NAD 27 coordinate for the DOT point and see if that doesn't fix the problem. Be sure you have NAD27 state plane for each point and not state plane for the NGS point and a surface coordinate for the DOT point. Don't know what kind of scale factor you have in the area but be aware of it for your calculations.?ÿ
If you can make that work, then convert the coordinates.?ÿ
The phrase adjusted to NAD 83 makes me suspect that it's an old NAD27 point that has never been re-occupied.?ÿ
I think any point with NGS NAD83 coordinates was at a minimum recalculated in the monster least squares fit of their raw observation data.
NGS doesn't use for their data sheets the simplified warping from 27 to 83 that CORPSCON does.
Two things come to my mind:
- NGS control points set by "traditional geodetic methods" were typically set to somewhere between 1:10,000 and 1: 100,000 precision. And they could be 20 miles apart. So a positional error of +/-5 feet or so, relative to a modern GPS determined position, may just be what it is.?ÿ ?ÿ
- Orthophotos aren't always perfect positionally. Depending on their purpose the ground control may have been very widely spaced. I've seen lots of such mapping that was only accurate to +/- 10' , 20' and more.?ÿ ?ÿ?ÿ
