« AnteriorContinuar »
A SIGNIFICANT SETI PROGRAM WITH SUBSTANTIAL POTENTIAL SECONDARY
BENEFITS CAN BE UNDERTAKEN WITH ONLY MODEST RESOURCES
A large, expensive system is not now needed for SETI. If we but equip existing radio telescopes with low-cost state-of-the-art receiving and data processing devices, we will have both the sensitivity to explore the vicinity of nearby stars for transmitters similar to Earth's, and to explore the entire Galaxy for more powerful signals, or for signals beamed at us. Such explora. tions, even should they yield negative results, would decrease our uncertainty concerning whether intelligent life transmitting powerful signals may lie beyond our solar system. At the very least, it would be of great interest and some importance either to know we have near neighbors, or to be reasonably confident no nearby transmitting civilizations exist. If, after we have made such modest searches, it seems important to us to embark upon a more ambitious SETI program, such as contemplated by the Cyclops study, the experience we will have gained will prove not only invaluable, but essential. Moreover, we expect to derive spin-off benefits of no small significance.
The arguments for electromagnetic waves as the communications medium seem compelling. The case for the nucrowave window seems very strong. The reasons for preferring the low end of the window are also strong, but not so strong that higher frequencies in the window should be ignored. The "water hole” between the H and OH lines is an especially attractive band that may be ideal for long range beacons. (See Sections II-4 and III-I.)
ETI signals, particularly those intended for detection by other searching societies, will probably be narrow in bandwidth compared with natural sources and may have monochromatic components which are as narrow as the interstellar medium permits. This increases their detectabil. ity at a given radiated power and distinguishes them from the natural background. The hardware needed for SETI therefore consists of an antenna or antenna system, low-noise wide-band receivers to cover the low-frequency end of the microwave window, means of resolving the received spectrum to a very high degree and means to search out and identify automatically any spectral anomalies.
Since halving the system noise temperature is equivalent to doubling the system sensitivity, it is important in SETI to have the lowest noise receivers that can be built. The background temperature in the preferred frequency region is only 6 K to 8 K (3 to 5 K in space) so every degree of reduction in receiver noise temperature is significant. The development of suitable low noise receivers represents a simple extension of present microwave technology and is not an expensive program. It would also benefit deep space communications and radio astronomy. (See Section III-5.)
To search for narrow band signals that may be anywhere in a wide frequency band and to do so in a reasonable time has been one of the major challenges of a SETI. In the Cyclops system concept the received signal was optically transformed into a high-resolution power spectrum. Since 1971 the growth of large-scale integrated circuit technology has been spectacular. It now appears possible to build, at reasonable cost, solid state fast Fourier analyzers capable of resolving the instantaneous bandwidth into at least a million channels on a real time basis. Development of such equipment is again a modest undertaking and the equipment would be very valuable for many other uses besides SETI. (For example see Section 11-5.)
To complete the data processing it is necessary to examine the power spectrum or a succession of samples of the power spectrum for any sort of significant pattern such as a sustained peak that may drift slowly in frequency, a regularly recurring peak, or arrays of regularly spaced peaks, to name but a few. The data rates are so great that this pattern recognition must be automated. The principal problems associated with the pattern recognition system are the amount of data storage needed and the identification of the types of patterns to be recognized. Only a few years ago these could have presented severe difficulties, but the solid state electronics revolution has so reduced the cost of memory, that prospective data processing costs appear to be relatively inexpensive.
It has been estimated that the development of the right data processing equipment would increase the capability of existing radio telescopes to detect ETI signals by about a thousandfold. This means that very significant searches can be made using existing antennas so equipped and it is recommended that the search begin in this way. The possibility of discovering some unknown type of natural source in this way must not be overlooked.
It is not feasible to search for all kinds of signals at all frequencies from all directions to the lowest flux levels at which a known signal of known frequency and direction of arrival can be detected. (See Sections Il-5 and III-2.) The more inclusive the search becomes in frequency on spatial direction, the more time is required, unless we sacrifice sensitivity. This is, of course, the reason for making use of all available a priori information and guesses as to preferred frequencies and likely directions of arrival. Many ingenious arguments have been offered for special frequencies and directions or even times; all can be given some weight as the search proceeds. On the other hand, every reduction in some dimension of the search is based on an assumption that may be wrong.
The strategy of searching nearby F, G, and K main sequence stars at ever increasing range seems very natural; the only life we know lives on a planet around a G2 dwarf star. This strategy takes us only as far into space as necessary to discover our nearest radiative neighbors around such stars. On the other hand, only slightly older cultures may be capable of radiating much more powerful signals, or they may know that life is to be found only around a few stars of a certain spectral class and age and may beam signals at these. As is true for stars, the nearest transmitters may not be the brightest. The strongest signals may come from advanced societies at great distances, whose transmitters may not even be near any stars.
For these reasons it is premature to adopt only one strategy to the exclusion of others. To cover a wide range of other possibilities it is recommended that in addition to a high sensitivity search of nearby stars, there also be a complete search of the sky to as low a flux level and over as wide a frequency band as practicable. (See Sections II-5 and III-3.)
To be significant, a full sky survey should be able to detect coherent radiation at a flux level one or iwo orders of magnitude below that provided by existing radio astronomy surveys. This turns out to be easier than one might expect. Although a sky survey as sensitive as ~3X1023 W/m2 has been made this has covered only ~2% of the sky. Another, covering most or the sky, has been made to a sensitivity of ~2X1020 W/m2. But in these, as apparently in all radio astronomy sky surveys, any coherent signals that might have been present were rejected as "interference.” Thus a complete sky survey using SETI data processing equipment to detect coherent signals at flux levels of ~1020 10 ~1024 W/m2 would be very significant. Existing antennas could be used to search the water hole to this level and the entire microwave window to as low as ~1023 W/m2 in a few years of observing time.
The target search of the nearer F, G, and K main sequence stars should be conducted using SETI hardware with existing antennas. This would permit detection of coherent signals at a Nux level as low as ~102? W/m2, or 109 10 10 times weaker than for the full sky search, assuming an observation time on the order of a half hour per star.
Both the sky survey and the targeted search could produce positive results, but even negative results will be of value since the upper limit flux levels that would result will be much lower than before. This could change our assessment of the capabilities of other intelligent life. The experience gained using SETI hardware in actual operation, with natural and man-made interference present, will affect the design of any future search strategies, and may lead to modifications of hardware, software, and search procedure. The searches we propose can be completed in approximately five years.
Planning a Dedicated Facility
SETI is more than a single effort. Like the exploration of the New World by our forefathers, like the present exploration of our solar system, it should be accomplished by many missions, each with some particular goal in mind. But there is a limit to the time that can be reasonably devoted to SETI from the facilities of radio astronomy or other services. To achieve the ultimate goals of SETI it will probably be important to have a dedicated SETI facility, the planning for which should begin now. This facility may never need to grow beyond a collecting area equivalent to one, or a few 100-m dishes. That will depend on future priorities, and on what we learn from the searches we immediately propose. The facility may be on the ground, or in space. (See Section II1-7.) We should, however, keep possible future needs in mind, and be prepared to build it whenever and wherever it appears appropriate.
Several ancillary programs should be initiated and pursued. These include protection of the water hole (1.400 to 1.727 GHz) against radio frequency interference (RFI) (see Sections II-4, III-8, and III-9), the detection of extrasolar planetary systems (see Section 111-3), the development of techniques for compiling extensive lists of target stars (see Section III-4), the study of alternative search strategies, and the continuing study of the cost effectiveness of space vs ground based systems.
In a resolution adopted at its fourth meeting the Science Workshop recommended that that international protection of the water hole against RFI be sought at the 1979 World Administrative Radio Conference. (See Section 111-9.) Navigational satellite systems are presently being planned that would destroy the usefulness of this prinie band of frequencies for SETI purposes. It is important to realize that for ground-based SETI systems such protection does not exclude all other services from the water hole, but only interfering ones such as satellites and nearby ground services. The RFI problem for space based SETI systems (especially systems in synchronous orbit) is more complex and probably more serious. Adequate shielding may be very expensive. It is not necessary that RFI protection of the water hole continue for all time. If no signals are found after a protracted sensitive search, the SETI priority may be relinquished.
The sine qua non of SETI is the plenitude of other planetary systems. While theoretical considerations suggest that planetary systems are common, it would be valuable to know how common and how their architecture varies with stellar class and multiplicity. Earlier astrometric telescopes and data reduction techniques could be improved to the point where the existence of near-by planets could be proved or disproved, but the effort might require two or three periods of a major planet, i.e., two or three decades. Preliminary calculations indicate that the direct observation of major planets around nearby stars should be possible with space telescopes of only modest size (on the order of one meter diameter). This could be accomplished by fitting the space telescope with a suitable filter or mask which greatly improves the contrast of a large planet with respect to the central star. Such an approach, if successful, would permit planets to be found in only two to three years after launch. This and other space techniques for direct planetary detection deserve active study and support. (See Section III-3.)
Present star catalogues list the coordinates of F, G, and K main sequence stars within only a few tens of light years of the Sun. If we ultimately carry on a search out to several hundred light years we will need to know the location of a thousand times as many target stars as are now listed. The problem of how best to conduct a whole sky star classification and cataloging program needs to be studied and, when solved, to be implemented. Since the compilation of such a target star data base must precede a major search, it is timely to begin the design study now. Both a greatly expanded catalogue of the solar neighborhood and knowledge about nearby planetary systems would be significant contributions to galactic and stellar astronomy as well as to SETI. (See Sections Il-6, III-4, and III-6.)
Although it is assumed that the searches performed in this program will be mainly for narrow band signals at the low end of the microwave window, other possibilities should not be ignored. Given a matched filter a series of pulses is just as easy to detect as a continuous wave (CW) signal of the same average power. The pulsed signal, however, introduces the new dimensions of pulse shape, repetition rate, and duty cycle. At this same time it is not clear that CW signals are more probable than pulses. Continuing study of these and other alternatives is indicated.
It will be seen that the program advocated above is of modest scale yet has potential for both SETI success and scientific contribution. Above all it serves as a logical introduction to the future but does not constitute a blank check commitment to a large expensive effort. The program is not a dead end nor is it open ended. It will be timely to consider whether to proceed with a larger scale program after this earlier effort has shown us more accurately what might be involved.