Description

In December of 2007, our group built a 44 input fast-readout spectrometer to search for energetic dispersed radio transients using the Allen Telescope Array. Each of our spectrometers processes a 209 MHz band, with 128 spectral channels and 0.625 millisecond integration time. For this experiment our 209 MHz band is centered at 1420 MHz.

Rather than using the ATA in the conventional interferometric mode, we point each of the 42 currently-installed dishes at different parts of the sky (in a horse-shoe pattern) to maximize our field of view, and allow the sky to drift through the antenna beams. Between February and April we logged close to 500 hours of observing time.

The two plots below represent frequency vs. time waterfalls for an 11 minute chunk of data from two of our spectrometers, connected to the x and y polarizations of ATA antenna 1E (x axis time labels are off by a factor 10^3).

Image:qg3XzlJN_1Ex.png|1Ex.png Image:qg3XzlJN_1Ey.png|1Ey.png

These next two plots represent similar waterfalls for another two spectrometers, connected to the x and y polarizations of ATA antenna 1C.

Image:qg3XzlJN_1Cx.png|1Cx.png Image:qg3XzlJN_1Cy.png|1Cy.png

These two detections (the bright regions in the plots) were separated between the two antennas by about 10 seconds.

The following five plots are zoomed-in versions of the 1Ey detection, at successively higher levels of zoom (x axis time labels correct)

Image:_S1wyVIG_00B.png|00B.png Image:_S1wyVIG_00B_long.png|00B_long.png Image:_S1wyVIG_00B_really_close.png|00B_really_close.png Image:_S1wyVIG_00B_zoom.png|00B_zoom.png Image:_S1wyVIG_00B_zoom_close.png|00B_zoom_close.png Image:_S1wyVIG_00B_zoom_more.png|00B_zoom_more.png

In the pointing configuration when these results were recorded (MJD ~54527.144), Antenna 1C was nearest the horizon, and antenna 1E was offset by 4.5 degrees.

The progenitor of the signal could be a satellite in LEO, but we aren’t completely sure. (assuming uniform circular motion and an angular speed of 4.5 deg/10 seconds).

=Table of (some) Detections=

Antenna Azimuth Elevation UT Date UT Time

ID (deg) (deg) (dd/mm/yyyy) (hh:mm:ss)

1E 37.5 24.0 02/03/2008 03:27:15

1C 34.5 20.6 02/03/2008 03:27:28

1E 37.5 24.0 09/03/2008 15:54:04

1C 34.5 20.6 09/03/2008 15:54:12

1E 37.5 24.0 10/03/2008 15:11:20

1C 34.5 20.6 10/03/2008 15:11:29

1E 37.5 24.0 07/04/2008 01:33:24

1C 34.5 20.6 07/04/2008 01:33:32

Other notes

The 42 ATA antennas were pointed in independent directions and autocorrelation spectra were recorded every 600 microseconds. Autocorrelation spectra had 128 channels covering 209 MHz bandwidth centered on 1420 MHz. Individual ATA antennas have a full width at half maximum power of 2.5 degrees at 1420 MHz.

The interference was detected on a number of occasions. In some cases, the interference was detected by two antennas with pointings separated by approximately 5 degrees, with a characteristic time delay of about 10 seconds. The power pattern in individual antennas appears to show rise and decay with a timescale also of about 10 seconds. The time in the Table is an estimate of the time at peak power, which corresponds to closest approach to the pointing azimuth and elevation.

eirp

$$
\eqalign{
  & SNR\~10 - 100  \cr
  & T_{sys} \~50K  \cr
  & BW\~1.63MHz  \cr
  & A_{effective}  = 0.6 \cdot 20m^2   \cr
  & r = range = \~1100km  \cr
  & EIRP\~{{kT_{sys} BW} \over {A_{effective} }} \cdot 4\pi r^2 \~0.1W \cr}
$$