Static GNSS Survey question

To be upfront, it's been at least a decade since I planned out baselines so take this advice with lbs of salt.

Is there any reason you can't place the 2nd receiver approximately in the middle of that 4 sided network??ÿ Aside from time/commute/cost reasons??ÿ P. 12 of the Leica 500 series quick start guide shows what I mean:

http://www.surveyequipment.com/PDFs/GPS500-General-Guide.pdf

I, personally, would always try and keep my baselines as close to the same length as possible.?ÿ A 2nd new?ÿ point, roughly, in the centre of the other 4 would help make "stronger" baselines.?ÿ Similar idea as a total station network.?ÿ A setup in the centre would have greater value than another side-by-side (you'd be mixing short and long backsites and foresites).

?ÿ

Side note--this reminds me of a debate we had at the old office in regards to baselines and gps checks.?ÿ For example:

50 km stretch of highway, 3 receivers.?ÿ Point A at the beginning, Point B at the mid point, and Point C at the end.

?ÿ

Person 1 Method:

-place receivers at A, B, C and log for minimum of 2 hours

-use receivers to infill from A to B, B to C

-create baselines from A to B to C

-create infill baselines

?ÿ

Person 2 Method

-use receivers to infill from A to B, B to C

-place receivers at A, B, C and log for minimum of 2 hours

-create infill baselines

-create baselines from A to B to C.?ÿ Use them as a check and/or add depending on how well they work.

?ÿ

As is likely evidently from the terminology I used, I preferred Person 1's method.?ÿ If the network itself had no errors when I "hung" it independent of published control, I was happy.?ÿ I'd then least squares adjust to published control.?ÿ However, I am open to debate as to which is better (or if another method is better than both).

Posted by

I, personally, would always try and keep my baselines as close to the same length as possible.?ÿ A 2nd new?ÿ point, roughly, in the centre of the other 4 would help make "stronger" baselines.?ÿ

If the gps network was combined with total station data (not in my case) in a least squares adjustment is the ts data not adding a similar length more local network line in a way?

Why limit yourself to 3 CORS? For the length of your observation times, 100s of kilometers are OK. Since you say you have a second receiver available and want to improve your proposed base, I suggest you occupy points within line of sight of your proposed control point and traverse between all. If you are limiting yourself to 6 hours, do a 3 hour on two adjacent points. If you are going for 4 hours, alternate 2 hour observations on 2 points, traverse and least squares adjust. Your 2 secondary points should form a strong triangle with your proposed base.

Back to my first question, use more than 3 CORS, in fact can you use 9? Submit OPUS-RS observations and note the 9 CORS used, then submit your longer observations to the 3 closest, the next 3 and then the last 3, paying close attention to the quality. Note that due to it's need for quality observations, OPUS -RS may not use CORS that are happily used by OPUS. Please note that in doing so you will have more office time than field time.

Paul in PA.

I don't see that you will gain anything unless you do a level run between the stations to improve elevation or a traverse to improve one axis of the horizontal error ellipse.

A second receiver will give a vector between the two stations, but if you have no other information to apply to the second station you can't use it to improve the solution at the first station. You know you are so high relative to "somewhere". So what?

Now if you do that run between stations, what error source will be different between GNSS sessions that the ground measurement will help?

Orbit and CORS errors are essentially the same so you get no help from the level run against them. At that spacing propagation errors will be nearly the same.

Multipath could be different, if it is significant, and the measurement on the ground could average down that contribution, especially if the remote point was better, but if you are choosing your bench mark point you will pick a good one to start with.

So I'd say you are much better off with added effort in processing more CORS as suggested above and/or running a session on a second day on your main point.

If you have two static receivers, just put one of them on a real benchmark?

The biggest question is, What is the benchmark going to be used for? I see surveyors get caught in the OPUS trap all the time when they really need to match local legacy information.

If you really need OPUS or CORS based 'absolute' height you won't get it with one observation. There are a lot of tricks to make it look like you did but most fail even a casual evaluation.

I am not in the US so Opus not an option here. I would have the option to add a few more CORS but are there not other factors such as Ocean tide loading past 100km? Local benchmarks not reliable?ÿ

In static observations, an additional gps receiver is always a good thing. Think of it as adding another CORS. Elevation from such a long baseline would still follow the manufacturer's specifications for vertical accuracy which is approximately 5mm +/-?ÿ 1ppm(?). So for a 20km baseline your elevation would be off by 2.5cm if my math is correct. If that's acceptable then both your receivers will be getting this accuracy even if you placed the 2nd receiver close to your observed point.

Posted by: @jt50

In static observations, an additional gps receiver is always a good thing. Think of it as adding another CORS.

How do you use the data to realize a benefit from the 2nd receiver??ÿ Are you assuming a level run between them?

In static, having two receivers on the ground within 5 miles and occupying control points gives double redundancy to your rover and helps to bring those error ellipses smaller and into your acceptable range.

Posted by: @a-harris

In static, having two receivers on the ground within 5 miles and occupying control points gives double redundancy to your rover and helps to bring those error ellipses smaller and into your acceptable range.

Yes, but the OP didn't seem to have any existing control points to work from and didn't mention running on the ground between his receivers.

Posted by: @totalsurv

I am not in the US so Opus not an option here. I would have the option to add a few more CORS but are there not other factors such as Ocean tide loading past 100km? Local benchmarks not reliable?ÿ

Are you in Canuckistan??ÿ If yes, do you pay to use any of the "big boy" networks??ÿ When I was using the SW Ontario Leica network you could call up Edgar and he would send you the RINEX file of the nearest stations if you asked for them.?ÿ I haven't used it for a while (10+ years) so I don't know if they have automated that now.?ÿ?ÿ

@bill93

when you use a nearby control point with known published coordinates, do you ask where it got its coordinates? Aren't all control points or CORS' coordinates referenced from some other CORS or higher class control point? So if the baseline from nearest CORS to the 2nd gps receiver is shorter than his 1st gps receiver then theoretically it should have a higher degree of accuracy compared to the longer gps?ÿ receiver point. So when post processing, I would first process the gps receiver with the shorter baselines from the CORS. Adjust its coordinates against all the CORS used. Then reprocess the longer gps receiver, this time with the 2nd gps point as an additional control point. Or use the 2nd gps receiver point as a sole control base for the 1st gps receiver point.

If we all had the time & resources to run a 1st order level run or traverse loop from a known control point then we would not be having this discussion. But gps is the next best thing available so might as well exploit it to its fullest potential.

First of all, when a benchmark is mentioned as the intended goal of a project my interpretation is that the primary goal is to be determine a height.

Secondly, as you are relying on an on-line processing tool, I assume you do not have software to do it yourself.

My comments are based on these assumptions.

1. Good ellipsoidal coordinates with respect to the CORS network can be obtained using long sessions with appropriate equipment, dual frequency receivers and antennas with known phase models, accurately measured antenna reference point (ARP) heights above the sought point are also critical.
2. While there are definite advantages to surrounding the project area with observations to points with known heights in order to better fit the new points with the existing vertical network,?ÿ this does not appear to be your situation.
3. The data reduction performed by the automated tools incorporate modeling beyond that of commercial packages.
4. An analysis of processing results (which differ between tools) is important. These tools can deliver a ƒ??badƒ? result (GIGO). Wrong inputs for antenna model and sub-type or ARP cannot be checked by these tools.

There are a number of tools providing positions worldwide beside OPUS like AUSPOS, GIPSY-OASIS, CSRS-PPP, SCOUT and others (see: https://www.gpsworld.com/a-comparison-of-free-gps-online-postprocessing-services/ ). Perhaps there is a more recent comparison?

What sort of height do you wish to obtain? Your observations submitted to any of the tools listed above will at a minimum provide an ITRF-compatible ellipsoid height. I have not checked the tools recently but be sure of the version of ITRF (currently ITRF 2014 epoch 2010 ).

Given an ellipsoid height (h) and wanting an orthometric height (H) one applies a model of the ellipsoid-geoid separation (N). Heights without error are related as follows: h - H - N = 0.?ÿ Remember also that these height surfaces are NOT parallel. In other words the difference in ellipsoid heights cannot be considered equivalent to the difference in orthometric heights.

There is a tool available on the UNAVCO site?ÿ international models using EGM08 and WGS84 (G873) coordinates (considered equivalent at the few cm level with ITRF 2014) at: https://www.unavco.org/software/geodetic-utilities/geoid-height-calculator/geoid-height-calculator.html

As for using another receiver near the primary point, you could validate the GPS derived orthometric heights by leveling between them.

?ÿ

Hope this contributes,

DMM

@jt50

I am always surprised when people quote these base error + length relative error specifications.

First of all, the best of these specification sheets include the proviso that the specifications are under ideal conditions. A good receiver is able to acquire and maintain signal lock and log all SV signals it is designed to do. 

Errors from sources like the troposphere (important in determining heights) are accounted for in data reduction software. The most critical elements in GNSS data reduction are frequency combinations, integer fixing, etc. are also handled in software.

If I take GPS data and process using OPUS first specifying close sites and then processing using distant sites, I do not see a correlation due to distance to the CORS. I see a correlation between length of session and precision.

Rather than trusting my small sample, see the paper "Accuracy of GPS-derived relative positions as a function of interstation distance and observing session duration" found here: https://geodesy.noaa.gov/CORS/Articles/Accuracy.pdf  

Expecting to achieve the accuracies claimed, especially over short baselines, is unwise.  

@geeoddmike

The intended goal is indeed to obtain an orthometric height in western Europe. I am using Magent Tools for post processing.

When you say the data reduction performed by the automated tools incorporate modeling beyond that of commercial packages, in what way?

@totalsurv

If you are using CORS, OPUS should be an option, or are you using the term CORS generally not specifically. In any case if post processing is your only option then my comments of occupy two close by points, with the second receiver still apply. Since you are setting control, multiple D&R from each of the three should be a part of your traverse, as you want observations that are truly worthy of Least Squares reduction.

Paul in PA

Posted by: @paul-in-pa

If you are using CORS, OPUS should be an option, or are you using the term CORS generally not specifically.

If you read all his posts in this thread you will find he is in Europe so must have the equivalent of US CORS to work with, but they won't be supported by OPUS.

If you just need an orthometric height in Western Europe, I would think that AusPOS would do you a good turn. Do you have good geoid coverage?

If the mark you are trying to evaluate has total open sky, then I think that you will not benefit from the 'extra' receiver.

However, if the mark has any obstructions above 25 deg elevation, then the 'extra' station will improve your results if 'extra' can be placed in the wide open.?ÿ

The general idea is that the very long vectors between the extra receiver and your benchmarks will evaluate well where the long vectors to the mark fail. But the very short vector between the extra and mark will evaluate with no issues.