Each grid point will have a convergence angle and each point will have a scale factor,?ÿ but they don't play into an adjustment of the measurements.
Note that while a CR adjustment operates on the coordinates it should be thought of as an adjustment of the measurements between the points.
convergence angles and why/how they??re needed to adjust what sounds like an open loop traverse?
The main reason convergence angles would be used in "open loop" or "link" traverses was because it was not uncommon to have two geodetic bearings at each end of the traverse (usually to a reference or azimuth mark, or from an astronomic observation thereof). It was necessary to apply the convergence angle at each end to find the grid bearings, so that when you went to perform your adjustment you were working out the math in a projected Cartesian coordinate system.
With a closed-loop traverse, strictly speaking it doesn't really matter whether you apply a convergence angle or not, because the starting and closing leg are one and the same.
@bill93 That's a great chart and coincides roughly with some testing I did a few years ago with a GPS + Glonass Javad Triumph-LS. I found that 4 minutes seemed to mark a point of diminishing returns in precision.?ÿ
I shared this on the Javad message board, and I'll share the highlights here. I've seen the epoch by epoch scatter plot of the vertical position of the RTK solution and over time it takes the form of a sine wave. I suspect the reason that 30 seconds tends to provide poorer precision than 5 seconds is because the 30 second observations may tend to be biased on one part of that wave (including only the peak or the valley of the wave and not some of both). I suspect this may be why the precision at 180 seconds is not below the straight line between 60 and 300 the way the horizontal precision is. Perhaps the 180 second average includes a full cycle or more (I'm referring to the cycle of vertical plot, not carrier phase cycles), and part of another, thereby partly biasing the solution to one or the other. Preferably, the average would include some n number of whole cycles of the vertical plot and not a fractional part.
Just a thought.
Also, multi-constellation RTK provides better precision faster in my experience. So this chart is somewhat antiquated at this point for most modern RTK systems, although the concept still applies.?ÿ