Setting Up Primary Control

The best way to combine GPS with traverse distance and angle (not bearing) data is a least squares program, giving realistic estimates for the accuracy of each type of measurement.

Does your CAD software have that capability?

@bill93

Thanks for your response, sadly I only have AutoCAD LT, so I assume not.?ÿ

I was thinking of just picking the two GPS points that are the furthest apart, draw a line between them, distribute the difference between the GPS distance and EDM distance to suit the EDM survey distance for these two points and align the EDM survey to these two points?

Very basic I know.

Posted by: @thomasd842

I?ÿshot?ÿeach point 5 times and took the average reading.

With GPS it matters how many epochs you include in each "shot". A single reading may be the average of several minutes of observations, or it may be an instantaneous position.?ÿ

I agree with Bill that Least Squares is the best way to combine surveys.

In CAD, you can make use of the ALIGN command to move one set of points onto the other.

?ÿ

The key here is what is your elevation nice sea level? What is the distance between your control points?

?ÿ

Posted by: @thomasd842

Bear in mind I am not a land surveyor and do not have the maths?ÿskills?ÿthat some of you may have.

Which leads to the question - Why are you doing the survey?

@steven-metelsky

It's not a particularly large site, it's 65m wide by 170m long, therefore the distance between point is slight bigger.?ÿ

The site is at +2.000m.

?ÿ

?ÿ

It's not a large site, 65m by 170m, it's a residential complex with a 15 storey tower, so not exactly complex, therefore Instead of getting in and paying a surveying contractor I thought I'd give it a bash.

Normally I would be given the external primary points and bring in and set up the secondary control, but I'd like to do it from scratch to learn.

The EDM traverse is now set and I am happy with that, I just want to know how best to position this network of points so the eastings and northings are accurate to the global grid.?ÿ

I was considering picking two corresponding GPS points and overlaying the EDM survey as close to the points as possible (CAD align function), so realistically with a GPS point tolerance of +-20mm I wouldn't lose to much sleep over it, but I thought I'd find out what the actual process is to do this to have a better understanding.?ÿ

@mark-mayer

Yep I plan on using the align function, but I just want to be sure I picking the two/three most accurate points to align to.

As the GPS points can be +-20mm out, I wanted to find out if there was simple way to set this up.

?ÿ

If I were doing this without least squares, I would first decide what datum and projection I wanted to work in. I note you are in England, and I am not familiar with what is commonly done there.

If you just want ground distances with bearings referenced to geodetic north, use software like our NGS program INVERSE (assuming your CAD doesn't deal with latitude and longitude) to find bearings and distances between the GPS points.

The GPS points are on the ellipsoid model of the earth and won't match ground distances at most locations. If you aren't at high elevations, that factor is usually less than 100 parts per million from unity and may be negligible on a small project.

I typed out a long description if what I would do, but then remembered you don't want to do a lot of math.

A simple and probably adequate way is to just put the two sets of values into CAD (different layers?) and visually rotate and slide one set until it lines up well with the other.

You should end up with one figure theoretically enclosing the other by a uniform distance (some parts per million). Due to measurement errors, one is likely cross in and out of the other.?ÿ

Consider whether the GPS or traverse is likely to be more accurate. If you think they are comparable, you could make intuitive adjustments to some points, but if a point is much further from its mate than at other corners, that's a flag that a re-measurement may be needed.

@bill93

Thanks Bill, I really appreciate your help.

Yes it's a site in the UK.

I'm a senior site engineer concentrating on structural package management, so the landing surveying aspects/maths are out of my league.

I'll probably go with overlaying the two surveys and tweaking to find the best fit.

Thanks again.?ÿ

@thomasd842

The problem is that you need to get the GPS points and the EDM points on the same coordinate system so that you can identify and evaluate the significance of discrepancies between the two. There's not an easy non-maths intensive way to do that.

Your suggested method assumes that one single EDM distance is correct. Thus, a lot of data (information) is discarded and there is no use of one system to check the other.

If there's not too much elevation change around the project, there is a way to approximate the EDM distances from the GPS distances.

For your longest distance, calculate the ratio: (EDM Distance)/(GPS Distance). Then multiply each GPS Distance by that ratio and compare it to its corresponding EDM Distance. The ratio is used as the reciprocal of the average project elevation factor. Multiplying it by the GPS (ellipsoid) Distance converts the GPS Distance to a Ground (EDM) Distance.

If you find a gross distance discrepancy, then you may have something wrong, or there may be too much elevation change for the average to be representative.

On the other hand, no gross distance discrepancies does not mean that the points are accurate. This is because an infinite number of point pairs can be the same distance apart.

We can go deeper, but it will require calculations.

?ÿ

?ÿ

@mathteacher

Thanks for your response.

There is a change in height, this is throwing my points out quite a bit I think.

I will give it one last go with shooting the points with the EDM with a random easting/nothing coordinate system and level (basically just turning the EDM on).

I'll install retro targets to reference back to and carry out the point survey from scratch ensuring everything is in the same coordinate system, therefore the EDM will do all the hard (maths) work.

I can shoot each point in both faces a number of times then average.

I can then plot this survey and overlay with the GPS and tweak.

If all else fails, I'll get in the professionals.?ÿ

?ÿ

Seems like we need more information here.?ÿ Is your client requesting these points be tied to a particular vertical datum?

Running a control network in the field requires experience, training and skill.

Performing analysis, quality control, and adjustment of a control network containing both GNSS and conventional observations takes a lot more of all three, and there are significantly more ways in which to screw things up.

If you don't have any experience with this sort of thing, I highly recommend handing it off to someone who does, or having them guide you through the process.

@rover83

Running good control is definitely a skill that takes a lot of learning, and being PLS licensed is no guarantee that a person has it.?ÿ

When you indicate that you took five (5) readings and took the average, do you mean that you setup over the sites and transcribed the autonomous readings on your display for subsequent determination of a simple mean?

Looking up your GPS, I could not find many details other than it is single frequency. Does it log GPS observations??ÿ

Perhaps someone on the board might recommend an on-line single-frequency processing tool?

?ÿ

Why are you doing this surveying if you are not a surveyor, by career or licensure??ÿ

You would need the right computer software, an experience in that software, to compute a least squares adjustment on the data.?ÿ

@jonathan50?ÿ ?ÿit's just control, it's not like it's a Boundary Survey or Stakeout.

@oldpacer I assume that the sarcasm font was lost in the translation.

I've never seen a project (boundary, engineering, you name it), that wasn't intrinsically constrained by some flavor of "Control Survey."?ÿ

????