GCPs collection workflow

Hi Rory,

What is you usual workflow for capturing GCPs then? Do you visit the site establish a base station, go home process the information then return to the site on another day with the processed coordinates of the base station, then use the rover to collect GCPs?

Thanks

Joe

Many people use NTRIP/CORS to get corrections to their rover instead of a traditional base. The advantage to this is that you only have to allow your rover to gather enough satellites and establish an RTK fix to the network station.

Personally I use a traditional base. One method is to set it up as soon as I get to a site and let it average in a coordinate while I get the rest of my gear ready which is usually about 15 minutes. Once the base has collected the point I document it and then enter it manually into Reachview. The downside to this is that the coordinate may not be perfect and match the CORS network which in my case doesn’t matter for the reason mentioned below.

A second method is using a known point for the site I am working on which allows me to enter a previously known coordinate that ties to the site control, but obviously this can’t happen if I am the first one on the site. If this is the case then the first traditional base/rover solution takes place.

I think the more important thing to keep in mind is that you want to be consistent so I will usually set a secondary base point in case my first gets disturbed or destroyed. I do this by collecting it with the rover for about 2 minutes with RTK corrections from my base on the first point. I work in construction so I usually have to reset a couple of the points in the middle of the job because there isn’t really a safe place to put them. My primary GCP’s are usually on the perimeter of the site so they are pretty safe. When I set new points I am still on the same coordinate basis whether I am on my primary or secondary base point. If you are on a large site it is also a good idea to set the secondary base point further away so you can reach other areas that the first position may have trouble maintaining an RTK fix.

3 Likes

Hi Joe,

With long baselines from our only OS Net reference station (>50km) we establish a base mark on the site using observations up to 16 hours, then use the mark as an RTK base for the SUA and the rover. With shorter baselines we now use a PPK workflow, you can read about it in this LinkedIn article: https://www.linkedin.com/pulse/post-processed-kinematic-gnss-mapping-rory-gillies/

The PPK workflow is actually more accurate as we are removing the chance of positioning errors of the RS2 over the base mark. Collecting checkpoints is also done using PPK, with the P4R we usually only collect five or six checkpoints that can be used as GCPs if required.

For sites we visit regularly then we almost always establish a base mark on the site, either from the first PPK mission or a separate visit, and mark permanent checkpoints using readily identifiable features on the site.

Cheers,

Rory

1 Like

While you will get some really qualified input on a question like that here, may I suggest you sketch out your own scenario you have to solve and what gear you are using?
Then we can be much more concrete in workflow advice.

1 Like

Hi Rory & Michael,

Thank you both for your reply’s.
I am trying to work out a workflow without ntrip for areas without signal.
I conducted a test using the base to average its coordinates and then the rover to measure a road.
when I post process the base against OS rinex information for my nearest OS base station I am then using the base solution to find an average coordinate for the base wich I then enter into RTKpost under the option tab positions for the base station (lat/long/height).
I then process the rinex file from the rover against this base position using kinematic positioning mode.

This goes through fine and the road looks good with mostly fixes from the rover, however when i pull the coordinates out using the time events from the survey csv and then convert that to OSGB36 my result is around 1m below the OS lidar for the area (and perhaps a little to the east).

I am sorry if i have typed this process out in a way that is hard to follow (I think i mite have lost myself along the way fiddling with the options in rtklib…)

If either of you could help shed any light that would be great.

Thanks

Joe

How long are your baselines and what gear are you using?

Hi Sorry Christian,

Using RS+ Base and M+ Rover.
Post processing local base using OSnet rinex from nearest base station
Post processing Rover using processed local base

thanks

Joe

Base line from local Base to OS Base around 25km
Rover to local Base no more than 80m

25 km on L1 is a little too long. You’ll see less of a fix-rate, and your stdev will be quite high.
I would suggest traversing the baseline with 1 or 2 intermediate steps, something like this: Translocating a Point w/ PPK Workflow

Or you could consider using a 24h hour obs for PPP.

3 Likes

Hi Christian,

Ok I think i understand, how would i go about doing that with the equipment I have?
Seems like i might be better off measuring GCPs in another way…

There are 2 main steps in this:

  • Obtaining the absolute basepoint for your RS+ to occupy, referenced to the CORS/NTRIP
  • Obtaining the GCP’s with the M+, referenced to the RS+.

That is, you need absolute accuracy anyhow?
If you “just” need relative accuracy, then your base position can be more sloppy.

2 Likes

Ah ok Christian,

If I just want relative accuracy as in using a local coordinate system can I just average my base position measure the gcps using the rover and then just post process the rover against the base without worrying about the OSnet base station?

What sort of relative accuracy can you hope to achieve using just the base and rover.

Thank you for your time talking me through this, you input is invaluable.

Yep. In that case you will 2.5 meter-ish absolute accuracy for the base position.

The numbers defined in the spec-sheet for the M+/RS+:

  • Static horizontal 5 mm + 1 ppm
  • Static vertical 10 mm + 2 ppm
  • Kinematic horizontal 7 mm + 1 ppm
  • Kinematic vertical 14 mm + 2 ppm
  • your mechanical/leveling tolerances.
2 Likes

Hi Christian,

Ah that not to bad, thanks for your input.
I am quite sure my process in rtklib is sound as i am getting good results using it.
I may look at getting an RS2 to use as a base in the future.

Kind Regards

Joe

2 Likes

Joe, what are you using to convert the ETRS89 coordinates to OSGB? The OS online converter or Grid Inquest has the correct geoids for the UK, I have found that using some software (FieldGenius for example) gives an incorrect Z reading for Shetland as it is using a different height reference. Is you height difference the same at each point? If that’s the case check your pole height and remember to add the Z offset in RTKPost settings, it doesn’t seem to read it from the RINEX header.

2 Likes

In my opinion your best bet is averaging the base, setting it manually and then setting up on the same point or secondary (if you set one) each time you return and if you need to transform use QGIS.

1 Like

Hi Rory,

I have just finished reading your write up on ppk GNSS Mapping on linkedin. I am using ETRS89 geodeic as the input on gridinquest II to convert to OSGB.

Thanks

Joe

This topic was automatically closed 100 days after the last reply. New replies are no longer allowed.