Our archaeology group has recently bought a ProMark 120 L1 rover with external aerial and GLONASS. We are using GNSS Solutions for post processing. We have a number of queries as we are climbing the rather steep learning curve.
We have found that both the manuals and other sources of information like YouTube videos do not appear to cover the basic use of a single rover operation. (Rover only no Base)
So we are struggling with the basics and would like advice on how to best use the kit.
Obviously we will be primarily using Stop and Go Kinematic, but we cannot understand if we ought to always initially do a Static BASE reading for 10-15 and indicate that this is a Control Point for the Stop and Go Kinematic readings that follow, or if that is not necessary.
What is the theory associated with control points during post processing.
In short how would you use the single rover kit for 10-30cm horizontal accuracy using Stop and Go Kinematics, re using a Static reading or Control Points/what to use as a Control Point etc?
Do we need to do anything special in GNSS Solutions re your recommendations?
Sorry for such a long question but many thanks in advance for your reply.
Our archaeology group is in Swaledale, North Yorkshire, UK see swaag.org
Thanks
Stephen
Just to clarify, do you have two units, a Base and Rover, or just a rover?
> Our archaeology group has recently bought a ProMark 120 L1 rover with external aerial and GLONASS. We are using GNSS Solutions for post processing. We have a number of queries as we are climbing the rather steep learning curve.
[...]
> In short how would you use the single rover kit for 10-30cm horizontal accuracy using Stop and Go Kinematics, re using a Static reading or Control Points/what to use as a Control Point etc?
Since you have only one receiver, should one assume that you're planning on using a second file from some GPS reference station to solve the vectors to your various points surveyed? How near is that other reference station(s) that you'll be using in post processing?
We will be using our nearest 1 or 2 Ground reference stations the nearest is 12Km the next is 40Km away.
Stephen
> We will be using our nearest 1 or 2 Ground reference stations the nearest is 12Km the next is 40Km away.
And you are planning to do only post-processed positioning, not real-time DGPS? What is the site like where you will be using the L1 rover to survey features, i.e. open field, woodland, or something intermediate as far as sky visibility goes?
It is an open rural aspect typically 14 sats in view. No realtime correction, all done by manual importing Base Station correction data into GNSS Solutions.
Uncorrected GNOSS data is typically 30 - 50 cm horizontal accuracy.
Thanks
Stephen
> It is an open rural aspect typically 14 sats in view. No realtime correction, all done by manual importing Base Station correction data into GNSS Solutions.
> Uncorrected GNOSS data is typically 30 - 50 cm horizontal accuracy.
On a site with the sort of sky visibility that you describe, if you keep the vector lengths close to 10km or less (12km would probably not be too much of a bother), I'd expect that you ought to be able to get 3cm accuracy or better in a post-processed solution from perhaps 15 seconds of data logged at 1-second or 5-second intervals.
The solution will require that you either (a) occupy an unknown point for sufficient time to resolve the interger ambiguities in post processing or (b) occupy a previously surveyed control point with an accurately determined position vector in relation to your reference station. Method (a) is the more fool-proof unless one has good confidence in the positional accuracy of the control points used for the second method.
If you are planning to spend a day or more mapping at a site, I would think about setting a few stable survey markers around the site, each with an i.d. number attached to it, and surveying them first by static methods. Then, you can initialize a stop-and-go kinematic segment very quickly on any of those markers as is necessary or convenient at any time during the work that follows.
One somewhat important factor is how active the ionosphere is at the time of the survey. Ideally, you want to be surveying with L1 gear when the ionosphere is relatively quiet if the separation distance between stations is more than 10km.
Here's a link to a space weather monitor. The planetary K-index, Kp, is a fairly good measure of how noisy the ionosphere should be. Kp = 3 or less is ideal.
Kent,
From your message it appears that you are confirming that when we get onto the open aspect survey site, that an initial static BASE reading for a sufficient time period at a random point is ideal, then to continue doing the required Stop and God Kinematic survey. Then in post processing import the static survey calling it a Control point.
Then include the Stop and Go survey data followed by the Reference Station data for the same time period. Then do the processing and adjustment in GNSS Solutions?
Can you confirm that please?
We are collecting data every second and making 5second Stop and Go kinematic readings currently.
Thanks
Stephen
To add a bit to the good information that Kent has offered, I believe that you need at least 5 epochs to process a solution in post processed kinematic applications, especially since you are using L1 only equipment.
So perhaps you'd do well to use a 1 or 2 second logging interval.
The other thing I would recommend, particularly where you intend to use a remote reference station for your base file, is that once in a while when you have good sky and a good dilution of precision numbers, set a reference point or occupy a sufficiently well-defined point that can be reobserved to reinitialize the system in case you lose satellite lock for whatever reason. Perhaps that is less likely with a dual-constellation receiver.
Jerry,
Many thanks. Indeed your last point is very timely as for the first time today we used a local BASE ref point that was logged for about 15 minutes before we started the main survey, and a poorly connected cable made it necessary to re-initialise by going back to that local ref point. I can see its value.
I have a GNSS Soultions query on today's data that I will submit next but it is late here in the UK so I will submit it tomorrow when I can reply to follow up queries.
Off to bed now - I will reply to others in the morning.
Cheers
Stephen
continue doing the required Stop and God Kinematic survey.
That must be an expensive option!
> From your message it appears that you are confirming that when we get onto the open aspect survey site, that an initial static BASE reading for a sufficient time period at a random point is ideal, then to continue doing the required Stop and Go Kinematic survey. Then in post processing import the static survey calling it a Control point.
> Then include the Stop and Go survey data followed by the Reference Station data for the same time period. Then do the processing and adjustment in GNSS Solutions?
In the interest of full disclosure, I should have pointed out that my remarks were not specific to the L1 Promark and GNSS Solutions. The survey controller must be asking you how you want to initialize the survey (i.e. resolve the unknown integer ambiguities) to kick things off and the best choices would be to (a) intialize on an unknown point by a static session or (b) to initialize on a known point with an accurate position vector from the reference station.
It may well be that the way the Promark survey controller is set up, you have to designate the static initialization as a control point, survey it, and then in post processing the Stop-and-Go session the software will automatically attempt to solve the integer ambiguites at that control point. I hope some Promark/GNSS Solutions user will shed light on that detail.
Learn The GNSS Solutions VRS Option
It utilizes multiple CORS stations to create a virtual reference observation inside your project area. From that VRS point you have very short bseline L1 vectors. Begin and end your observations with longer than a short static observation and you should be very pleased with your results.
Paul in PA
