a2048: GPS l3 remains on

04apr18

Links to sections

Dynamic spectra of the 9 scans
gps L3 power vs time
Satellite locations
Summary

Project a2048 took data on 04apr18 01:37 to 05:13 UTC using alfa and the mock spectrometers. They reported GPS l3 interference for most of the run.

Their setup was:

alfa and the mock spectrometers covering 1220 to 1520 Mhz
1 second integrations
9 x 1200 sec scans. ( i ignored the cal scans)

Dynamic spectra for the 9 data scans

Scan start time 04apr18 UTC	file number	scan_id	image
01:37:46	20180403.xx.00500	809300062	img
02:01:42	20180403.xx.00800	809300065	img
02:25:38	20180403.xx.01100	809300068	img
02:49:34	20180403.xx.01400	809300071	img
03:13:30	20180403.xx.01700	809300074	img
03:41:23	20180403.xx.02000	809300077	img
04:05:19	20180404.xx.00200	809400003	img
04:29:15	20180404.xx.00500	809400006	img
04:53:11	20180404.xx.00800	809400009	img

The table below identifies the different rfi bands

freq	source
1381.05	GPS L3
1349.6,1344.4,1257.6,1252.5	FAA radar
1294.5,1279.5,1243.5,1228.5	Punta salinas radar
1288,1278	galileo satellite E6 band
1268	beidou satellite
1246	glonass?

Total power GPS L3 versus time

The average power 1381.05 +/- .5 Mhz was computed for each 1 second spectral sample.
The data was normalized to Tsys when the rfi was off.

the plot shows the GPS L3 frequency power vs time (.ps) (.pdf)

The data is plotted vs time (HH:MM:SS UTC).
The colors are the 9 different 1200 second scans.
the satellite transmission was off when the level dropped to 1.
this shows that the satellite transmitted almost continuously for the 3+ hours.

Location of satellites

The GPS satellites are in 1/2 sidereal orbits (every l1.x hours). (not sure about galileo).
Plots were made of the az,za (relative to AO) for each satellite of the GPS and galileo constellations. The
pass selected was the closest to the observation times (00:00 01:00 ast 04apr18).
The right side of the plot labels the satellite and the AST date/time for the closest approach of the pass.

GPS constellation (.ps) (.pdf)
glonass constellation (.ps) (.pdf)

These are source azimuths (directions of the satellite).
the source azimuth,za for the telescope varied from az:210 to 250, za:10 to 18 degrees.

The final plots shows the GPS satellites that were above the horizon during the observation (.ps) (.pdf)

Since the GPS L3 rfi was present for the entire time, I only kept GPS satellites that were above the horizon for the entire 3 hours.
Page 1: Angular distance between GPS satellite and arecibo beam vs time.

None of the satellites were particularly close.
GPS19 was the closest to the beam (27 deg) at the start of the observations.

Page 2: az,za of the satellites during the observation period.

the red + lines at the bottom show the az,za (source) for the telescope

Page 3: AO azimuth of the satellites during the observation.

dashed red lines are perpendicular to the sides of the triangle.
It is easier for a source perpendicular to a triangle edge to scatter into the dish and up into the dome.
None of the satellites stayed close to 60,180,300 degrees for the entire time.

Summary:

GPS l3 remained on for 3+hours on 04apr18 utc 01:37 to 04:53
this was seen in the alfa data taken at this time.
The assumption was that a single satellite was causing the problem:

There had been no notification of a test.
If it is a single satellite, it is probably one of the 7 listed.

none of the GPS satellites were within 27 degrees of the aoBeam
We also saw galileo interfering at the E6 frequencies : 1278, 1288
The 1268 interference was probably the chinese beidou satellite,.
To eliminate satellites from consideration, we need more data during another pass.

I checked the a2048 data on 01apr18 and 06apr18, but no constant GPS L3 transmission was seen.

processing: x101/180404/gpsrfi .pro

page up

home_~phil