[time-nuts] Effects of data length and observation intervals on RINEX results

[Mark Sims]

I had a 16 hour long / 1 second interval RINEX file produced from a Trimble Acutime GG receiver and decimated it several ways to compare how the results changed.   

The Acutime GG is pretty much the worst performing device in my collection... it is L1 pseudorange only,  it is a small sealed unit with a built-in patch antenna,  and it was sitting 8" above the ground in a horrible multi-path infested position.  

The first way the data was decimated was to reduce the 16 hours of 1 second interval data to 1/5/10/30 second intervals.  The lat/lon/alt error ellipses were:
1 second:  188/175/488  mm
5 second:   259/241/673 mm
10 second: 297/277/773 mm
30 second: 371/346/966 mm

The second way it was decimated was to take the first 2/4/8/16 hours of 1 second intervals:
2 hours:   628/667/1756 mm
4 hours:   439/415/1022 mm
8 hours:   288/270/738 mm
16 hours: 192/178/496 mm

The results with Ublox LEA4/5/6/8T (carrier phase and pseudorange) receivers with 16 hours of 1 second intervals were around 125/130/250 mm.  With 2 hours of data and/or 10 second intervals the errors were around half of what the Acutime GG produced.