Hi All, Im new here and at all this so apologies if I have done this wrong.
We currently do 3d mapping with our process being create our map in the office, then go and fly it (Inspire 2, Phantom 4 Pro, Map pilot and DJI GS pro), which gives us geo referenced x,y,z JPGs. We then choose 4-5 points on site for ground control points, and create a SSF file from Trimble Geo6000XH with external antenna and Terrasynch CM version, which we then take back and post process with Pathfinder Office, by getting the latest CORS/RINEX data and getting a report files on the 4-5 points normal 80-95% 0-5cm.
My question is this, if we changed our workflow to the Emlid RS, would we need one or two. Would one work in the field by collecting the 4-5 points, then using this information coupled with RINEX/CORS data back in the office to post process, so we have a single device, or do we need two RS’s - one at a “new” base station on site GCP CONTROL for example, then a second RS to take to points 1-4(5), then when back at office compile report from the unit taking the points, unit acting as base, and CORS/RINIX?
This all sounds very confusing to me, but then my background is emergency services not GIS and Im struggling.
Any help or advice would be gratefully appreciated. My requirement is to create geo-accurate GCPs of 1-2cm ideally, which we can then use in Photoscan Pro and others to geo-rectify images for processing.
Many thanks - and sorry if this is a silly question
Using a pair of Reach RS will give you the best precision because they are two identical receivers with two identical antennas. However, the next best is to use a CORS as your base station.
Let us say that you can create a map using GCPs and get 3cm precision. Also, let us say that experiments at your project location prove that precision with a pair of Reach RS is about 1cm and that accuracy with your CORS corrections is 5cm.
So, if you use a single Reach RS with CORS corrections, each GCP is then accurate to 5cm, and any point on your map will be 5 + 3 = 8cm precision, and the map will be placed in the world with: 5 + 3 = 8cm accuracy
And if you use a pair of Reach RS with CORS corrections (only for the base location), the base is accurate to 5cm, and each GCP is precise within 1cm relative to the base. Any point on your map will be 1 + 3 = 4cm precision, and the map will be placed in the world with: 5 + 3 = 8cm accuracy
That is a bit over-simplified, but you get the idea. Having a second Reach RS doesn’t align your map to the world any better, but it makes your map more precise.
I am fortunate enough to have access to the Australian AusCORS network. This gives me far better accuracy and reliability than a pair of reach RS’s (when within 10km). Over longer baselines I will use a long static observation to set up a local base. Not sure what would happen if I had access to a networked CORS correction, i.e. one where I send my position and I get an interpolated correction set for my location. Nice summary BIDE,
Irony is I will normally get a stable FIX solutoin from the longer baseline quicker than the shorter local baseline.
There a whole bunch of reasons why long baselines with L1 only is not a good idea. BUT. Most can be mitigated by a bit of effort, and the best thing is most of the effort is not human effort (which costs money) but equipment and number-crunching effort which can occur without supervision, just takes longer.
Even if you use an L1/L2/L5 survey grade GPS long baselines need care and degrade with distance.
All those extra frequencies do is allow you (in realtime) to estimate the propagation impacts of the particles between your receiver and the satellites. In Post processing you cab download better estimates of those effects and use them. You still will have errors and bias that you cannot detect or quantify unless…
YOU simulate a networked solution by calculating your position using multiple reference stations at different angles and distances to your new point and compare the results. Recently I set up a new base with a 45km baseline (as an experiment) I was fortunate enough to be 5km from another Auscors reference. I compared the two results and was still sub centimetre in agreement. This was with a 20min observation period.
If I was doing a 100km baseline I would observe for 24 hours (its only data after all) and process at 0.1Hz. For any and all reference stations within 200KM and compare results. I would also ensure that I had all the RTKLIB boxes filled with respect to precise ephemerides and ionospheric correction.
HOWEVER. Why not just find the nearest control point and occupy it with a reach and use your own shortter baseline?
Unfortunately, some work Ill be doing doesn’t have local control points - some will and for those Im fine! Most people will use rtk feeds and not worry about post processing; however for me and uav mapping Im of the opinion that post processing is the way to go.