Hi Greg,
many thanks for your very detailed answer.
I will dig deeper into the docs.
On 17.12.25 01:10, Greg Troxel wrote:
The first is that you should make sure you are using the same datum.
Positions from gpsd, if using EGNOS (guessing you are in .at), are in
the frame of EGNOS. That's a harder question than you might realize.
See page 20 and onwards:
https://www.euspa.europa.eu/sites/default/files/brochure_os_2017_v6.pdf
If not using EGNOS, then it's WGS84.
Yes I am in Austria and it is EGNOS. But where do I choose it ?
With CSRS-PPP, they have mulitple offerings and you didn't specify.
After that, if you are using a dynamic datum (Egnos TRF, ITRFnn, WGS84),
or a static datum on the wrong plate (NAD83 in EU), then you need to
bring positions to a common epoch to compare them.
Uploading the file I used the button ITRF because this sounds for me logical
* The epoch will be the same as the GPS data.
* A UTM zone will be calculated from the longitude.
The most interesting question is "what is the distance?" The maximum is about
80 cm ( 2,6 ft )
I don't find that super interesting because I don't expect accurate
answers from averaging GPS.
You are right, it's completely irrelevant what’s the position of the
antenna on my house.
But you know, we all are striving for maxims: higher, deeper, faster,
further
And in this case "closer to a narrower area"
In general averaging can do a good job, but not always.
For example I am at 48 north. Therefore the satellites are visible from
the south up to the zenith and also east and west.
But in average all satellites signals have a longer way through
atmosphere which could ( I don't know ) distort the result but cannot be
corrected by averaging. For an observer near equator it's maybe not such
an issue.
Therefore I think the CSRS-PPP service brings hopefully the better result.
I do find the rapid->final CSRS coordinate shift interesting (guessing
at pairs, would be cool to draw lines). And the clustering of the 4 PPP
results.
Maybe I find the time for this.
It would be nice to also show some kind of error bounds on each. I
suspect that showing error bounds will point out that the error bounds
are not credible.
I was also thinking about this drawing ellipses instead of points.
Kind regards
Hans
