AST 309
part 2:
Extraterrestrial Life
The Search For Extraterrestrial Intelligence
The Drake Equation:
N = N*
Stars? Planets? Habitable
Origin Complex Intelligence, Lifetime
of life? life?
technology? of civilization
Another way to see what the
Drake equation represents:
What is the likelihood or
probability that the
sequence of events shown
to the right will occur,
ending up with the peculiar
group of phenomena listed
at the bottom branching?
And are there many other
possible branchings that
we haven’t thought of, or
are incapable (for, say,
biological reasons) of
Intelligence, technology
Intelligence—If we want to communicate with aliens, they must have
something like what we usually call “intelligence.”
What does this mean? Are there different “types”? What about other
Other cultures? Especially oral tradition cultures.
Why did humans evolve such big brains?
Why think that extraterrestrials would share our forms of cognition?
Factor in Drake equation:
What is probability that life elsewhere develops “intelligence”? Call it fI .
Communication, technology—If we want signals, the aliens must have
some kind of system of representation, like a language. How likely? Other
forms? Musical? Unimaginable? And no matter what kind of language they
have, they have to be technological civilizations if they are to send signals
across the Galaxy.
We could enter these topics as two more probabilities, fC fT .
N(planets with life that also developed intelligence and technology) =
N* fpl nhab fL fC fT
Lifetime of the communicating phase
We left out one crucial factor: The fraction of the Galaxy’s age
T that a civilization is in the technologically communicating
phase. We know the age of our Galaxy is T ~ 10 billion years.
Consider our chances if civilizations only last for 100 years…
Average length of time spent in the communicating
phase L. Sometimes referred to as lifetime of a technological
civilization. Call it L. Our final Drake equation is then:
N = N* fpl nhab fL fC fT (L/age of Galaxy)
This represents the number of communicating civilizations in our Galaxy.
We have been “this way” for only ~ 100 yr. If this were L, then N (number
civilizations for communication in our Galaxy) would be around 1.
Even if N ~ 100, there is no hope for two way communication, and
probably very difficult to locate them among a 100 billion possibilities. And
too far for biosignatures.
In order for our Drake equation “N” to come out reasonably
large (say a million), the average lifetime L must be extremely
something like 100,000 to a million years.
 This means we can expect any nearby civilizations to have had a
long lifetime, i.e. be very advanced. I think this is the only thing we can
SETI web sites
Ongoing SETI searches:
SETI Institute/Allen Telescope Array
Project SERENDIP, UC Berkeley
[email protected]
Sourthern SERENDIP, Univ. Western Syndey
SETI Italia
Optical SETI at
Amateur SETI:
Project BAMBI
SETI League
Project ARGUS
SETI is concerned with searches for signals from extraterrestrial civilizations, not
spectral “biomarkers” we discussed earlier in course, or actual travel to other star
 Except for 1974 signal to globular cluster M13 (thousands of light years away), we
only try for reception, no transmission. (“What if they are all listening?”)
 “Signals” could be intentional (they are trying for contact) or nonintentional (we
eavesdrop, one way or another). A few unintentional candidates listed on next slides,
but we are most concerned with intentional signals, and with establishing a two-way
 Remember: Distance to even the nearest 1000 or so stars ~ 50 light years, and we
expect only a tiny fraction of them to have life, let alone intelligent communicating life,
so we are necessarily asking whether to undertake a search with a guaranteed very
low probability of success, and, even if successful, communications will involve
decades or even centuries.
You can see why funding for SETI is sparse!
Four types of nonintentional types of signals:
 Leakage radiation from radio, TV, or other radio broadcasts.
Earth as example: Many TV stations broadcasting different shows, or same shows
at different times, radio waves emanating from Earth have always been incoherent,
completely scrambled. It is not true that alien ETI could be seeing our early TV shows!
 Alien communications, e.g. between home planet and colonies.
We would have to be almost exactly in the line of sight between home planet
and colony, and guess the frequency range.
Seems very unlikely.
 Dyson spheres -- hypothetical constructs built by advanced civilizations in order to
collect nearly all the energy of their parent star.
Spherical shell at same distance from star as the home planet. The intercepted energy is
channeled to planet. But shell is heated by the incident radiation: What will its temperature
be? At what wavelength would you conduct a search for Dyson spheres?
 Products of technological activity – e.g. gamma rays from their (hypothetical)
fusion propulsion systems, … CFC molecules from their air conditioners…
TV leakage radiation
World TV spectrum
World TV power vs. time
Geographical distribution of TV stations
Geographical distribution
of TV stations
As Earth rotates, this “pattern of
populated areas” is the only evidence
for TV broadcasts
How would an alien civilization try to communicate
across many light years of space?
The only thing that is almost certain is that they will use photons—fastest and
cheapest way to transmit information that exists (as far as we know).
Even though photon signals are the only choice we can think of, that still leaves
many considerations that we need to guess about:
 Where to point our telescopes? What kind of stars should we point our
telescopes toward? Or would it be better to survey the whole sky?
 Wavelength: What wavelength region should we expect is optimum for sending
interstellar signals? Radio? Optical? Other?
 Bandwidth: What range of wavelengths? Broadband or narrow-band? WHY?
 Recognition and Interpretation: How would a message, or some sign that it is
not a natural phenomenon, be distinguished, and how would a meaningful
message be encoded?
We’ll discuss each of these in turn.
a. Sky survey. Survey entire sky with telescope’s “beam” – this might
involve millions of directions for typical radio telescopes. If you want to finish
in your lifetime, you could only spend a very brief time on each direction, so
could only detect very strong signals. But at least you won’t miss any of
And the method doesn’t make any assumptions about what the most likely
stars are
for signal reception.
 This is a low-sensitivity method, but complete for strong signals.
b. Targeted search. Point at the nearest (less than about 50 to 1000 l.y.)
stars roughly like the sun and cooler (recall conditions for habitable
planets). Could detect weaker signals, i.e. would have higher
But you will only cover a tiny fraction of the sky.
 This is a high-sensitivity method, but severely incomplete.
Most current searches have shifted to a sky survey mode (a) However
plans change rapidly—the Allen Telescope Array (largest current
project) combines both approaches.
What frequency should be used to listen or
send interstellar messages?
From the Earth’s surface, most radiation
is blocked by the atmosphere.
The exceptions are optical (visual)
and radio photons.
Earth’s atmosphere also blocks out most of the infrared part of the spectrum due to
vapor in our atmosphere. From the highest mountains or a jet plane, the infrared is
barely accessible, but not for the continuous kinds of surveys we have in mind.
Note that if we could do such a survey from Earth orbit (expensive), or, if we only
had about $100 billion dollars so that we could build a facility for SETI on the far
side of the moon (“Project Cyclops”), our considerations might be different.
Why have most SETI searches concentrated on radio wavelengths instead of
A single amazingly influential paper by Cocconi and Morrison (1960 Nature) set the
Their arguments for radio SETI are on next slide.
Reasons for radio SETI:
1. Interstellar dust selectively blocks shorter wavelengths (higher
frequencies) strongly suggests we use the IR or radio parts
of the spectrum (long wavelengths or small frequencies).
But IR is dominated by Earth’s atmospheric molecular
emission if search from surface, so that leaves radio.
Notice that radio SETI allows reception from the entire galaxy, but
optical isn’t that bad, since we can see stars out to ~ 1 kpc.
Besides, most radio searches are concentrating on nearby
stars anyway. (We don’t want 1000-year “conversations.”)
2. Radio photons are cheaper to send than optical photons
(because energies are ~ 100,000 times smaller for radio).
3. The main consideration is noise: Here “noise” means anything
that is not an alien signal--any kind of interference.
We should listen (or send) where the noise is minimized, so that
we can
recognize the (probably weak) signal. Noise is minimum in a
region of
the radio part of the spectrum. This is summarized in a classic
Allen Telescope Array,
shown on next slide.
SETI Institute, N. Calif.
[email protected], Phoenix
Alien signaling: Choosing a wavelength range that
minimizes “noise” -- anything that is not an alien signal
Avoid very low frequencies
(wavelengths too large), because
synchrotron radiation from
supernova remnants dominates
there. (Far left in figure)
Avoid frequencies higher than about
10 GHz because of H2O and O2
emission from Earth’s atmosphere.
Cosmic microwave background
radiation sets “floor” at intermediate
frequencies, and that is where
the noise is minimum, and where we
should search.
A message will arrive in a narrow wavelength band or bands, not spread over the whole
1-10 GHz region. There are 10 billion 1 Hz bands in this range. How to decide which
ones to pick? First, must understand bandwidth.
The importance of bandwidth
Basic idea:
Can pack more power in a narrow frequency range (narrowband signal) than
spreading out over a large range (broadband signal).
So can distinguish a narrowband signal from the background more easily.
Think of the everyday radio analogy again, and it should be clear!
[email protected]: Each vertical “band” is a 10 kHz “slice” of the 2.5 MHz wide SERENDIP data.
There are 250 such “slices.” But search is for signals much narrower than these bands.
Which frequency?
If it is true that narrow-band signal is the only sensible approach, how will we
decide which
band to use?
Suggested “beacon frequencies” (or “hailing frequencies” or “magic frequencies”):
HI (neutral hydrogen) 21cm (wavelength) line? The frequency is 1420 Megahertz =
Natural, abundant, but lots of interference by interstellar gas. (Latter has apparently
been forgotten.)
OH line at 1.7 GHz? H + OH = H2O, so maybe region between these two
--> “the waterhole”. Alien civilizations will know that these two lines are from
the dissociation products of water, whatever they call H and OH. Not taken too seriously,
But convenient range to strive for. (Allen Telescope Array uses this range, and more.)
•Some frequency based on combinations of fundamental constants of nature? (e.g.
speed of light, Planck’s constant, …) The combination can be expressed without
referring to “our” units (e.g. meters)
•”Intergalactic” frequency standard based on temperature of cosmic background
•Many others have been suggested. Too many! None in use today.
SETI projects: partial historical list
Ozma (1960)
In 1960, radioastronomer Frank D. Drake, then at the National Radio Astronomy Observatory
(NRAO) in Green Bank, West Virginia, carried out humanity's first attempt to detect interstellar
radio transmissions. The stars chosen by Drake for the first SETI search were Tau Ceti and
Epsilon Eridani.
From April to July 1960, for six hours a day, Project Ozma's 85-foot NRAO radio telescope was
tuned to the 21-centimeter emission (1420 MHz) coming from cold hydrogen gas in interstellar
space. A single 100 Hz channel receiver scanned 400 kHz of bandwidth. The astronomers
scanned the tapes for a repeated series of uniformly patterned pulses that would indicate an
intelligent message or a series of prime numbers such as 1, 2, 3, 5 or 7. With the exception of an
early false alarm caused by a secret military experiment, the only sound that came from the
loudspeaker was static and no meaningful bumps superimposed themselves on the formless
wiggles on the recording paper.
SETI projects: partial historical list
Arecibo: message sent to M13
• Globular cluster ~25,000 lyr away
• Message written by Drake, Sagan, et
• Frequency: 2380 MHz
• 1679 binary digits
SETI projects: partial historical list
Ohio State University SETI
With the Big Ear fixed radiotelescope
(used Earth’s rotation to scan the sky)
August 15, 1977: detected a strong narrow-band signal for 72 secs:
• Duration consistent with extraterrestrial source
• Strength: 30 times above background noise
• Frequency: 1420 MHz (neutral H)
• Bandwidth: <10 kHz
SETI projects: partial historical list
The location of the signal in the constellation
Sagittarius, near the Chi Sagittarii star group.
RA= 19h25m31s ± 10s
or 19h28m22s ± 10s
declination= −26°57′ ± 20′
There were ~50 follow-up searches performed in this area, but nada
=> Origin of the Wow! signal is still undetermined
SETI projects: partial historical list
NASA asks for SETI proposal, astronomers propose “Project Cyclops”, 1000
100 meter radio telescopes on back side of moon, costing $10 billion (1970s).
NASA asks for more moderate plan, planning for next ~ 17 years.
Ohio State SETI: 1977-1997 (replaced by golf course).
Best known for the “wow” signal.
Harvard, Paul Horowitz and Project META (millions of bands in frequency), Project BETA (billion bands in
frequency). Horowitz and Sagan 1993 Astrophysical Journal summarize results. One of first SETI
papers in refereed journal.
Harvard and Horowitz now converted to Optical SETI, largest in world.
UC Berkeley’s Project SERENDIP. Since 1977! Part of data analyzed by 5 million home computers
through [email protected]
Dec. 1991. NASA funds $100 million SETI effort (“MOP”). Detailed plan for combined targeted and sky
survey searches. 1993: Funding removed by senate amendment
Project Phoenix (SETI Institute) rises from the ashes. Piggy-backs off various radio telescopes, mainly
2001: Paul Allan and others fund the Allen Telescope Array, 350 6meter telescopes. 42 complete by
Oct. 2007.
One of oldest operational SETI searches--since 1979, UC Berkeley.
1997--installed as piggyback at Arecibo radio observatory (picture below), largest
single-dish radio telescope in world (but can only point in one direction).
SERENDIP = Search for Extraterrestrial Radio Emissions from Nearby Developed
Intelligence Populations. SERENDIP IV is the fourth instrument of the project, collects
data by 'piggybacking' on top of the Arecibo radio telescope.
SERENDIP IV instrument is basically a 200 billion operations per second supercomputer
that scans 168 million narrow (0.6 Hz) channels every 1.7 seconds for signals that
are significantly 'louder' than the background static.
Some of its data is analyzed through [email protected] for desktop
computers--so far millions of users, largest distributed computing
project in world, led to ~ 100 other distributed computing projects,
e.g. [email protected], [email protected], [email protected], …
(details on next slide)
The Arecibo radio telescope in Puerto Rico,
used by both SETI Insitute for Project
and by UC Berkeley for their SERENDIP IV.
[email protected]: Now searching for pulses
August 2008: [email protected] switches to search for pulsed radio signals.
Observations are from SERENDIP piggy-backed on radio telescope at
Arecibo that is built into a mountain. This dish only points in one direction
as sky drifts across this direction--the drift takes about 12 seconds for a
given point in the sky.
[email protected] searches for signals that rise and fall in 12 seconds--any object
will do this, but most will be broad-band sources (top image).
Narrow search by requiring narrowband signal (2nd image).
Will check for several different bandwidths.
Information in image? Search for pulsed signal (3rd image).
If from planetary system, should also be Doppler shifting
(“chirped” signal), as in 4th image.
Home computers look for various combinations of frequencies, bandwidths,
and chirp rates. See if you can understand why the white “thing” in the
illustration below might be a signal…
[email protected] screensaver
Can you see the
alien signal??
[email protected] and distributed computing
By 2007 over 40 projects had joined the BOING distributed computing family,
using the software provided by [email protected]
Protein folding: [email protected], [email protected], [email protected], [email protected] privately run mathematical project that searches for very large prime
numbers and has already found more than 100 new primes.
[email protected] is based at the University of Wisconsin in Milwaukee and searches
for pulsars in the sky based on data from the gravitational wave detectors LIGO and GEO. Oxford Univ. UK high-profile climate simulator.
Jan 2005: First paper in Nature, 2,570 simulations of Earth.
By 2007: 50,000 simulations. Goal is several million, to explore
23 parameters of the climate model.
Allen Telescope Array (ATA)
Eventually 350 6 meter antennas, equivalent to 100 meter single dish.
42 dishes saw “first light” in Oct 2007.
Entered hibernation due to funding problems this year!
=> Maybe restart in Sep 2011 due to private donations > $200,000
Unique features:
 Large field of view, so can scan sky faster in survey mode.
 Large range of frequencies (1-10 GHz for targeted search, five times range
of Phoenix), and small bandwidth (~ 1Hz), using more than a billion
 Finally offers SETI 24/7 monitoring
(Phoenix had Arecibo for only about 3 weeks per year 1998-2004)
 Targeted search: Survey 106 stars with good sensitivity between 1 and 10
GHz for weak non-natural transmitters.
 Sky Survey ~ monitors inner Galactic plane in “water hole” range 14201720 MHz for very strong non-natural transmitters.
Optical SETI (OSETI)
When faced with the question “What kind of signals would alien civilizations transmit,
the traditional answer has been: Continuous narrow-band radio transmissions
An alternative: Maybe they would send distinct broad-band pulses. They would stand out
against background noise not because they are precisely centered on a particular
wavelength, but because they are very short and punctuated bursts of energy—unlike
most other natural phenomena. This is the world of Optical SETI, which searches for
signal in the visible (or infrared in future) light range, usually looking for nanosecond pulsed
laser radiation.
Radio 21cm line has huge noise problem with interstellar gas.
Laser-like signals are tightly beamed, so can be sent over very large distances (no loss due to inversesquare law). With current equipment can send out pulsed laser beam 5000 times brighter than the Sun
Unidirectional--can pinpoint direction with high precision. Higher frequency --can encode more
There are two OSETI programs at UC Berkeley, and now ongoing or about-tolaunch
OSETI searches
projects:at Harvard, Lick Observatory, and Columbus Observatory.
Berkeley: Monitors 2500 stars as part of exoplanet search. Searches for ultra-narrow band signals.
Harvard: since 1998, using 61 inch telescope. Nearly 100,000 observations.
Dec. 2000: new 72 inch telescope dedicated to an all-sky survey. Can detect nanosecond (billionth
of a second) pulses and cover entire sky in 200 nights.
Lick Observatory Targeted search began in 2000. First results published in Stone et al. 2005
Astrobiology. 14 candidate events
Encoding a message
SETI researchers focus on a signal anyone could comprehend. Not clear this is sensible!
 It is very sensible to expect digital, binary, not analogue signal.
 How to encode a picture into a string of binary (0,1) signals?
The simplest and most efficient way to encode a message (we think) is binary code. Use only 2 characters,
e.g. a 1 and a 0, or a + and a - , or "on" and "off", ... Each 1 or 0 (or whatever) is a "bit". Then the message
can just be sent as a series of pulses.
Expect the message to be a two-dimensional picture that is encoded in a one-dimensional
binary string that factors into prime numbers.
e.g. 551 = 29 x 19 (or 19 x 29); 1679 = 23 x 73 (the 1974 Arecibo transmission).
Example: We receive signal 1111100000101011010110101.
This factors into 5 x 5, giving a picture of the greek letter "pi".
Or try the letter "E", etc.
But why would ETI send out signals that anyone could decode? Perhaps they send out signals which
could be understood only by others who are already "at the same level" as they are.
What would be a difficult signal for us to recognize?
Maybe the test would be to recognize some sort of "meaning" in the message. (Think about musical
signals. At present, there is no viable theory of musical meaning in music analysis, philosophy, cognitive
science, pattern recognition, or any other field that has approached the problem.)
Deeper questions: Will symbolic communication systems be universal among intelligent creatures? Is
“grammaticity” hard-wired into our brains? Another example of single mutation?
The Eerie Silence…..
 SETI has been on-going for about 50 years and NOTHING has been
 Conclusion: are we alone? Are there no ETIs in our galaxy?
 No, not yet. But they are probably no nearby ETIs similar to us, nor a
nearby strong radio beacon intended to notify us of ETI’s presence.
There are many stars not observed yet, so many frequencies not
explored and many other possible ways for ETIs to communicate.
What IF The Eerie Silence is broken!
What then?
There is actually a so-called “SETI Post Detection Task Group” (a permanent
study group of the International Academy of Astronautics). It’s only a think
tank, no legal status, no governmental involvement.
In case a signal is detected the task group would advise to perform the
following steps:
1. Verify authenticity of ETI signal
2. Contact the International Astronomical Union => which will contact UN.
3. inform the government of the country (countries) where radio telescope is
4. Public announcement
Such a discovery will have unprecedented and enormous scientific,
social, cultural, political and spiritual consequences for humanity.
Nobody knows what will happen (depends on nature of signal, e.g.