Hello, I was wondering about suggestions on processing monitoring data collected using RTK. We have several monitoring points at a site in which we collected 8 observations per point throughout the day. Does if it makes sense to do an adjustment with this data using the standard deviations for each point provided in LGO (Leica Geo Office)? We have StarNet available to do adjustments. I am only somewhat familiar with doing adjustments to static data. Since there is only one baseline in RTK I am a little confused on how to make sense of doing an adjustment on this type of data. I am pretty new to this so any suggestions would be greatly appreciated!!
Thank you for your time
Jen
If you put the coordinates into Star*Net with the standard deviations of each, it will do a weighted average for you.
Hi Jen,
Assuming that I understand your question (which might be assuming a lot), I would ASSUME that you have 8 RTK vectors from your BASE to the ROVER separated by [lets say] 1 hour between "shots" (observations).
Now further assuming that each observation was the result of multiple epoch solutions (1 minute, 2 minute, whatever), AND (this is important in MY opinion), EACH shot is based on a separate INITIALIZATION between shots, then;
In your data collector there should be:
A dx/dy/dz vector from the BASE to the ROVER
A 3x3 covariance matrix (or rms + correlations) for each vector
And usually a bunch of stuff like Number of birds, PDOP, HDOP, VDOP, GDOP, number of epochs, etc.
ASSUMING that everything (above) looks OKAY, then you could simply create input records for each shot that contains the dx,dy,dz, xx,xy,yy,xz,yz,zz data for each vector (BASE to ROVER(1), BASE to ROVER(2) ...), and dump that into your Least Squares Software. Now the format of THAT file will vary somewhat based on your LS Software, but in any case, the resulting Coordinate Estimate (and residuals) "should" tell the tale, and be the best guess as to the Coordinate of the Rover based on ALL 8 shots, and the associated statistical data returned by the RTK Engine (receiver). All things being equal (yeah...) the answer should be very close to the arithmetic mean of the 8 coordinates returned by the data collector in the first place.
At least that's [pretty much] how I deal with multiple RTK observations on important points.
Loyal
Jen,
The first thing, I would caution you on using RTK for monitoring for anything other than substantial movement. I have used it on major landslide project looking for significant movement on the slope when it is full of workers. By significant I'm talking over a tenth or two signaling the half of the mountain was going to join the other half a thousand feet below.
That being said, it sounds like you are using Leica gear. Our 1200 systems provide a weighted average, as described by Bill, on the collector. You can also adjust the points holding the base a fixed in LGO. For single base observations I have never notice a difference with the weighted mean but it does give you a little more information to reference in a report.
Have you thought about two base station locations?
The way I do this is by exporting a gfile from my receiver and importing the gfile into Star*Net. I have to edit the resulting .gps file in order to get the point names right, but that's pretty easy to do.
Starnet will test the vectors against each other as well as the estimates you provide. The advantage over LGO is the results will be expressed in standard terms rather than overly optimistic proprietary language. Even the LGO manual states the 'QC' number should be about 2.5 times larger...
