A Comparison of SETI Programs

A Figure of Merit

Plotted here on the X-axis on a log scale is the equivalent isotropic radiated power (EIRP) of a just-detectable ETI transmitter, versus, again on a log scale, a measure of the "volume" of signal space searched to this detectability level. This "volume" is defined as the product of the number of stars similar to our sun that could be detected if they had transmitters broadcasting with powers greater than or equal to the given EIRP and the number of MHz of spectrum searched for each of these stars. This is the best way we could think of to put sky-survey and targeted searches on the same scale. It does no good to point at a target star if the system is not sensitive enough to detect its transmitters (if any). Likewise, a particular beam on the sky is uninteresting unless it contains a star within detection range. For both targeted searches and sky surveys we are assuming the probability of success is proportional to the span of spectrum searched. No consideration is given here to the different types of signal each search can detect. The number of stars classified as "good targets" has here been estimated as one half of the total number of main sequence stars of types F, G, and K, using counts from the RGO catalog for stars within 50 ly and the mean local density of such stars for greater ranges. The factor of one half makes a rough allowance for the fact that some stars will be unsuitable because of their orbits around companions or their youth. Optical observations were begun under the NASA HRMS Project to identify the specific best targets; these are continuing at present.

Recently Completed Searches

These curves compare three recent SETI programs. The Phoenix Parkes curve represents the search conducted by Project Phoenix for sixteen weeks in 1995 at the Parkes Observatory in Australia. The other two curves show the most sensitive prior SETI observations. The Planetary Society's META observations at Harvard are described in Horowitz and Sagan, Astrophysical Journal 415, 1993. The UC Berkeley SERENDIP III observations are described in Donnelly, C., Bowyer, S., Werthimer, D. and Malina, R.F., 1994, SERENDIP: The U.C. Berkeley SETI Project, Center for EUV Astrophysics, Publication no. 557. Both of these systems are in the process of being upgraded.

Planned Searches

The above chart compares SERENDIP IV, BETA (successor to META), and the planned Phoenix observation program. These comparisons are based on the best available information as these programs get under way. Also shown, for comparison, are the EIRP values of the two most powerful transmitters on the Earth (from Billingham and Tarter, Acta Astronautica, 1992): the Ballistic Missile Early Warning System (BMEWS) and the Planetary Radar at Arecibo. (The latter achieves its extraordinary EIRP both from a very powerful transmitter and from the extremely high gain afforded by the Arecibo antenna; this gain, alas, also means that it is less likely that one would "see" such a narrowly beamed transmission by chance.) Commercial TV stations have EIRPs of less than 0.01 GW.

For the highest power transmitters, the merit of Project Phoenix rises less steeply than the others because it is searching to greater distances than the other searches. At distances of more than 1000 ly the flatness of the disk of the galaxy reduces the number of background stars in the direction of some of the target stars.

Other Figures of Merit?

Other "Figures of Merit," making different assumptions about the distribution of ETI signals, could be assigned to a given set of SETI observations. As an example, if one assumed the only detectable signals would be deliberate beacons and further assumed that such beacons would be at a guessable frequency, for example that of the positronium 1(3S1) -> 1(1S0) transition (Kardashev, Nature 278, 1979), then the corresponding search "volume" would not have a term proportional to the frequency range of the search; indeed, were this the case, none of the sets of observations shown would have any "merit" since none cover the frequency of this particular transition.

The design of the Phoenix search system assumed that we can guess what kinds of stars would be most likely to harbor technological life, and guess limits to transmitter powers, but that we cannot guess the frequency of the transmissions very well. The design of the META system, by contrast, assumed that we cannot guess where technologies may develop, nor what limits may exist on the power of the transmitters, but that we could guess that the ETIs will (sometimes) build very powerful beacons at a guessable frequency (one or two times the frequency of the hyperfine transition of neutral Hydrogen).

John Dreher
System Scientist, Project Phoenix
dreher@astro.berkeley.edu

More information about Project Phoenix may be obtained from:
phoenix_info@seti-inst.edu
SETI Institute - 2035 Landings Drive - Mountain View, CA (415) 961-6633