We are not a survey company, therefore we are going to avoid digging through the published network of coordinates (known monuments) throughout our province. We also do not have an RTK (NTRIP) subscription yet. Located in CANADA. We have two Reach RS2 receivers.
However, we need at least 0.50 - 1.00m absolute precision.
Looking through the docs, there’s some options we have.
Now, here’s my questions & options (preferably help me out with Q2!)
Average single position - I’m assuming that this is an autonomous position. Accuracy of 2.5m. Does the accuracy increase as the occupation time increases? How are we looking at say, 2 hours of static observation?
Post-Processed Kinematic - this requires no NTRIP subscription, correct?
From what I understand, basically we take the two RINEX files (one from the rover, one from the base) and process this in software after the survey is completed? Then the corrections are applied?
In the notes: “15min on PC after log from reference station available, usually posted hourly”. What reference station are they referring to? Something from NRCAN that is not PPP?
Also, what software do you recommend? RTKLib? Emlid Studio?
Precise Point Positioning - I have used NRCAN’s PPP service before. The only downside is the time constraint.
Thanks guys. Appreciate it and happy to become apart of the Emlid community.
No NTRIP connection, Correct. For PPK, you need to locate a CORS station within a reasonable baseline distance to your project that offers Rinex observation logs. Emlid Studio is being developed for the processing.
Given this scenario:
Area with no cell reception, no known points and no nearby CORS reference station. Absolute neck of the woods. You have two Reach RS2s. What is the best way to get an accurate base position? Allow the base to log for hours? Assuming open sky where the base is occupied.
I’ll defer to some of the more experienced on here. But, if you need absolute accuracy, I think you’ll need to go with a PPP workflow. In the middle of nowhere, why the requirement for absolute accuracy vs. relative accuracy on your project?
So the workflow would look something like this in that scenario:
Set up base on unknown point (min 4 hours). Set the base to logging following the steps in Emlid doc
Collect rover points. Establish link between rover and base over LoRa
Pack up. Export data.
Preferably wait a day so the PPP service has better corrections. Submit the base RINEX logs to PPP
Find the nearest CORS station and bring in the RINEX file for that day
(not too sure here - does Emlid studio do all of this?). Bring in the corrected PPP coordinate and modify the old base point to the new corrected point.
The rover points either shift automatically or (depending on software) I have to manually adjust the .csv export based on the deltas X, Y and Z from the new PPP point.
Bring in that CORS (nav or obs?) file and do another process with these extra corrections. Repeat the previous step of adjusting rover points.
A bit jumbled on step 6. Is the order correct? Not sure if I should process the base with CORS before submitting to PPP or after.
I’m assuming the rover data is processed after the base - because rover is all relative to the base. These steps, from what I understand, are just to get an accurate absolute coordinate for my base.
I’ve done these steps more or less in Leica Infinity before, and it was fairly straightforward. Unfortunately our budget does not allow me to purchase the software.
I just want to add some points in addition to the true suggestions of Christian and Dave
You can discard the 5th step of your workflow. PPP service is indeed an appropriate way to obtain absolute accuracy of the point at the neck of the woods. And it doesn’t require any logs from CORS to calculate it. You can use updated coordinates from PPP to calculate base shift and apply it to the points collected by the rover.
Regarding the 6th step, this base shift has to be applied manually in the CSV file. Otherwise, you can use Emlid Studio for the workflow called Stop & Go.
In this case, you don’t need an RTK link between base and rover. You only need to record logs on both receivers and collect points in ReachView 3 in SINGLE mode. You use the base log for the PPP service and add these updated coordinates in Emlid Studio. The rover log and CSV file are used to calculate the corrected coordinates of points. Please note that this workflow allows you to get only geographic coordinates, not projected ones. You can read more about it in our docs.
In this case, you don’t need an RTK link between base and rover.
Is that because it is unnecessary (i.e. the final result will be the exact same with RTK or without) since this is technically a PPK survey? And if the two receivers have no real-time connection between them, the baseline is good up to 100km?
Say the base and rover were connected through RTK. The base does not have great absolute accuracy because of no NTRIP and its over an unknown point. Wouldn’t it make more sense to submit the base RINEX to PPP for processing, then adjust the rover coordinates after? The rover coordinates already have great relative accuracy to the base.
With the Stop & Go method, it seems more processing is done to get the same result as above. Correct me if I am wrong, I’m still trying to wrap my head around it all!
Please note that this workflow allows you to get only geographic coordinates, not projected ones
That’s fine. I’m assuming with the corrected csv file from Stop & Go, they are in a WSG84 system (EPSG 4326). I can then just do a batch EPSG transformation to my desired projected EPSG? I have a Python script that does this.
Well, that is not incorrect at least. Post Processing offers a very high degree of flexibility, as you can decide things like SNR and horizon cut-off thresholds, exclude separate sats, tweak start/stop, exclude/include constellations etc.
You can also utilize backwards processing, and use more compute-heavy algorithms for ambiguity-resolution etc, etc.
In general, I always log raw-data, so if RTK-data shows to be bad or marginal, I at least have the option of trying to post-process for a decent solution.