ALFALFA Survey - Examples of commonly seen RFI     Last update Wed Sept 28, 2011

Useful links

Bad Beam Bogus Birdies: It's not really RFI.

More often that we would like, one or more of the ALFA beams is dead. The power to its amplifier should be turned off. But the way the data-taking schemes works, an array of data is being recorded for the spectrum. Bogus data are bogus data, with unpredictable results for autoscaling. Most of the quick-and-dirty data monitoring tools ("alfadatawin", "quicklook") assume all spectra are good and use autoscaling to display. So, if you have bogus data, you may wind up with totally bogus autoscaling

The image to the right shows an example of the "corplot" display (same as "quicklook") which you get if you use the IDL recipe (or "quicklook" which looks at Record 0 only) to look at a particular record. In this example, Beam 4A is dead: Note that the beams here are displayed with labels 1 to 7 not 0 to 6, so the dead beam is the red line with spikes in the Beam mis-labeled "5". This is not RFI; notice also the difference in the vertical scale of the plot for this beam relative to the others.

Know what this is and then ignore it. Be sure to switch the display in the waterfalls plots to ignore the bad amplifier; in this case, set it to display Beam "B", not "A" or "both". Also, if you can't seem to get the waterfall plots to display correctly, switch the "alfadatawin" display to view the "spectral plots" and set the y-scale to "constant". Take a look also at the power level monitors in "alfaobswin". If the power is steady, then there is no RFI -- you are just seeing the bogus data from the dead beam.

Click here to enlarge.

The San Juan FAA radar

As noted on Phil's list, one of the frequencies (1350 MHz) used by the San Juan airport FAA radar lies within the ALFALFA bandpass, so it almost always shows up. It is pulsed, polarized, azimuth-dependent and not picked up equally by all beams. When it is strong, harmonics may show up at 1380 and (sometimes) 1410 MHz.

The image to the right shows an example of the main transmitter plus the (much weaker but still visible) 1380 MHz harmonic.

There is not much we can do about the radar; it's needed and entirely legal. Please record at the bottom of the log file whether the FAA radar is on (or off!).

Click here to enlarge.

The Puntas Salinas radar

As noted on Phil's list, the Puerto Rico Air National Guard station at Puntas Salinas has a landing system which operates at 1367 and 1382 MHz. It is not supposed to be turned on when we are observing (i.e. its use is not supposed to interfere with observations at AO). See also this link.

The image to the right shows an example of the Puntas Salinas transmissions (also visible are the FAA 1350 transmitter plus the 1380 MHz harmonic, discussed above).

If you see the dual frequency signature of Puntas Salinas, ask the telescope operator to call the folks at the air station to tell them to turn off their equipment. Be insistent, if necessary; call yourself even. Also, if it persists, file an rfi report by emailing "". Include in the message the AST time, frequency, description of the RFI, and please cc Martha (haynes_astro.cornell). Also, of course, include a note in the observing log file.

Click here to enlarge.

The Global Positioning Satellite Nuclear Event Detector

On occasion, you may see one or more short bursts of rfi, lasting from 30 to 180 seconds, centered at 1381 MHz and extending over 1-2 MHz. These bursts arise from the NUclear DETonation detection (NUDET) system aboard the Global Positioning System (GPS) satellites. This system was established in the late 1970's as the US was planning to sign the Comprehensive Test Ban Treaty giving up the right to conduct tests of nuclear weapons. The NUDET system was designed to allow the US to verify that other nations are in compliance with the treaty. See this link.

The image to the right shows an example of a GPS/NUDET burst (along with the FAA radar at 1350 and the mysterious new 1375 MHz birdie whose identity is not yet known.) Notice that GPS/NUDET is polarized and not identical in all beams. If you are curious, Phil has a set of IDL routines which predict the passage of satellites over AO.

Click here to enlarge. You might want to compare that with this normal display without GPS

If you notice up to a few, short (<300 sec) GPS/NUDET bursts, make a note(s) in the log file.

The use of the 1381 MHz frequency for GPS/NUDET was carefully negotiated back in the 1980's to minimize impact on radio astronomy while still insuring national security (No kidding: this discussion including folks whose offices were in the White House...!). As a result, there are limits to the burst duration and total accumulated time per day, month etc that the NUDET system can transmit for system tests under "normal" circumstances. Once in a while, the Air Force notifies the observatory of planned systems tests which may exceed the normal limits; we post received NUDET test notifications. If you are observing when there is not a posted alert and the duration of the GPS/NUDET seems excessive, file an rfi report by emailing "". Include in the message the AST time, frequency, description of the RFI, and please cc Martha (haynes_astro.cornell).

Fortunately, GPS/NUDET bursts results from system tests and do not mean that an above-ground nuclear event has taken place so you don't need to dive under a desk. (That is a joke to us old folks.)

Original page created and maintained members of the Cornell ExtraGalactic Group and the ALFALFA team
Last modified: Wed Sep 28 16:05:34 EDT 2011 by martha