PROJECT PHOENIX: THE AUSTRALIAN DEPLOYMENT
Jill Tarter for the Phoenix Team
Copyright: Stuart Kingsley
From February 2 until June 6 of 1995, the Phoenix Team conducted SETI observations using the 64 m radio telescope at Parkes and a remotely operated 22 m antenna at Connabarabran. The dual polarization observations covered the frequency range from 1.2 to 3 GHz using a single wideband receiver and two feeds built by CSIRO to support this project. The two antennas simultaneously observed a target list of 202 solar-type stars located at declinations south of -35 deg. Individual observations lasted up to 276 seconds and examined 20 MHz of the spectrum with resolutions as fine as 1 Hz using hardware pattern detectors to search for narrowband, continuous or pulsed signals whose frequency might be slowly changing with time.
The data were analyzed, and candidate signals were identified in near real-time (before the end of the next data acquisition cycle). Those candidate signals not matched against an on-line RFI database were automatically reobserved with finer resolution by another set of detectors and followed in phase in order to permit a pseudo-interferometric measurement between the two telescopes. This two-stage approach (detection on the 64 m antenna and immediate interferometric follow up of candidates) was part of a pipelined observational sequence and proved to be extremely effective and efficient method of discriminating against RFI. Detection thresholds were set to produce a few candidate signals per observation, yet for more than 23,000 completed observations, the programmed sequence had to be interrupted fewer than 100 times to move the antennas off source for further verification procedures. In each case the candidate signals were found to be of our own technological making.
The Phoenix observations in Australia failed to detect ETI signals, but they also left no mysterious or unexplained signals hanging around. The deployment was a logistic and technological success, and reaffirms our opinion that one telescope is not enough.