The limits of SETI
We haven’t found aliens yet, but how far have we looked?
Suppose there were another planet just like Earth. How far away could we detect it?
Seen from the moon, the Earth would still be large in the sky, about 4 times as big as the moon looks from Earth:
But the Earth gets really small, the further away you get:
As the Voyager 1 probe left the solar system, operators instructed it to turn around and use its camera to look back at the Earth. Here’s what it saw:
Earth is that one pale blue dot in the image, and the colored bands are lens flares from the sun.
Voyager also took a wider angle image, including the sun, which puts this image in context a little bit better:
When that photo was taken, Voyager was about 4 times as far from the sun as Saturn is. The closest star is 4 light years away, or 6,000 times further from the sun as Voyager was. It would be very hard to take a picture of Earth from another star system — it would be very faint and also very close to a much brighter star.
None of the telescopes we have, on the ground or in space, are capable of taking a picture of an Earth like planet at a distance of dozens of light years away.
We have found planets around other stars, though, using a variety of other methods.
Finding exoplanets
We’ve already found almost 6,000 planets outside of our solar system.
We’ve even been able to take pictures of a few of those planets, but only if they’re both larger than Jupiter and also orbiting far away from a star.
For most planets, we need to rely on more subtle methods. In some cases, we can see that a star is wobbling slightly from the gravitational pull of the planet. In other cases, we can detect when a planet crosses in front of a star — the light reaching us from that star dims slightly.
Both of those methods are limited in what they can find. When we look at how stars wobble, we mostly find large planets, particularly those very close to the star — hot Jupiters. When exoplanets were first discovered, this was the only method which could find them. That did not mean that most planets are hot Jupiters, though, it just meant that we couldn’t find the smaller ones. Smaller planets don’t perturb their star’s orbit enough to detect.
We later learned how to find smaller planets by watching them transit across a star. That only works if another solar system just happens to be tilted in the right way, so the planets cross the star as seen from Earth. This technique will miss most solar systems because they aren’t tilted in just the right direction. But it can find lots of planets, and it can give us statistical clues about what sizes of planets exist.
The Kepler mission launched a satellite looking for planets transiting across stars. It mostly found planets between 1 and 4 times as large as the Earth, orbiting close to their stars:
The most common planet found was between 1 and 4 times the size of the Earth, orbiting close to the star, with a year less than 100 days long.
That’s an interesting conclusion because none of the planets in our solar system are like that. We have smaller than Earth planets with a short year, and we have larger than Earth planets with a long year.
But, again, the data is biased by what the telescope can find. The Kepler mission confirmed that hot Jupiters are rare, those are only found around 2% of stars. But even the smaller planets it found were all still close to the star:
The smallest planets found were half the size of the Earth, but those could only be found if they were very close to a star (with a year less than 20 days long). Earth sized planets could be found with a year up to about 50 days (closer to the star than Mercury is). And large planets could not be found unless their year is only about 300 days.
The Kepler mission would not be able to find another planet just like Earth, with a 365 day long year.
A newer satellite (TESS) has also been finding exoplanets for a few years. It’s a bit less sensitive than Kepler, but it will search most of the sky. It still won’t be able to find an Earth sized planet with a 365 day orbit, but it will be able to find larger than Earth rocky planets, Earth sized planets orbiting close to their star, or Earth sized planets orbiting close to a red dwarf star. Some of those might be habitable but we don’t really know because we don’t know the rules for life forming.
We’ve found enough planets to have a rough guess as to how common habitable planets are: one paper says that 1 in 5 sun like stars have a habitable planet. More specifically, they mean a planet between 1–2 times the size of the earth, receiving between 1/4 and 4 times the amount of sunlight that Earth does.
That gives about 10 billion chances for life, in our galaxy.
But actually finding those 10 billion planets will be hard — if they are just like Earth, we could not take a picture of them or find them with any of our planet hunting telescopes.
Finding life on those planets is going to be an even harder task.
Searching for oxygen
The Earth’s atmosphere did not contain much oxygen until photosynthetic cells evolved. It then took another billion years before the oxygen started to build up and accumulate in the atmosphere.
Today, you can tell that the Earth is alive just from looking at the atmosphere, which is roughly 79% nitrogen and 21% oxygen.
Finding oxygen in another planet’s atmosphere would be a good sign of life.
We have not yet found oxygen in another planet’s atmosphere. Our technology is barely at the level to look, though. We’ve surveyed maybe 100 exoplanets’ atmospheres, and most of those were hot Jupiters.
We can survey a planet’s atmosphere as it transits across a star. We’ve surveyed a few small planets that way, looking for things like hydrogen, water vapor, or carbon dioxide.
So far we haven’t found oxygen, but the number of habitable planets surveyed is very low (I think it’s only in the single digits).
The James Webb telescope can look more closely at some of the planets found by TESS. As we discussed, none of these planets will be exactly Earth’s size with Earth’s year length, they will be bigger and closer. If life is common in the universe, though, some of them might still be habitable. We might be able to see oxygen in those planet’s atmospheres, and we could discover that in the next 10 years or so. That would be the first sign we see of (simple) extra-terrestrial life. The number of habitable candidate planets these two telescopes can look at won’t be very high, though — one paper says that TESS will only find dozens of planets that can be analyzed and only several that might be in the habitable range.
If life is less common, or it needs exactly Earth’s size and distance, then we won’t find life until we build better telescopes that can directly take a picture of an Earth sized planet from dozens of light years away. As I mentioned before, that’s a very difficult task. One paper describes the challenge:
Any Earth-like exoplanets within dozens of light years are about as faint as the faintest galaxies ever observed by the Hubble Space Telescope, but first, to detect biosignatures, we have to divide the light into individual wavelengths to detect spectra; hence, we will ultimately need telescopes larger than the Hubble. Second, even more challenging is that these exoplanets are adjacent to a parent star that is up to 10 billion times brighter than the planet itself. The challenge of direct imaging of an Earth analog is similar to the search for a firefly in the glare of a searchlight when the firefly and searchlight are 2,500 miles distant (the separation from the east coast to the west coast of the United States).
The technology to do this is complicated. You need to block out the light from the star to see the planet. It might be possible to block the light inside the telescope, with a very sensitive design. If not, you need to send out a second spacecraft, flying 40,000 miles away from the first telescope, that will make a tiny eclipse to cover the star.
I don’t think either of these things are going to be built soon.
If life is common, we’ll find oxygen with the combination of the TESS planet search and the James Webb telescope. If it’s not, it could take a decade or more for technology to get better, so we can see Earth sized planets at just the right distance from some nearby stars and look at their atmospheres.
In either case, finding oxygen would only prove the existence of simple life, and it might still be disputed — there is one theory as to how oxygen could show up without life, on certain rocky planets.
Until then, we have another way to look for advanced life: searching for radio waves.
Searching for radio waves
We could not photograph another Earth with our current telescopes. To have any chance to discover that planet, its solar system would have to be tilted in just the right direction so that we can see it cross the sun. Even in that case, our current scopes still wouldn’t be able to see it, unless that other planet was in an orbit closer to its star than the Earth’s actual orbit.
Could radio astronomy do better?
Suppose that another Earth had a civilization just like ours, and it was emitting the same radio waves. Could we detect those radio waves?
Radio astronomy has a few advantages. The first is that the telescopes are much larger. The largest optical telescope has a 10 meter wide mirror. Radio telescopes can be much bigger than that. The former Arecibo observatory in Puerto Rico was 300 meters across, built into a natural depression:
There’s now a larger 500 meter radio telescope in China.
It’s also possible to use multiple radio dishes to collect one signal.
There are plans to build a bigger collection of dishes, the square kilometer array, which should have a total area 5 times as large as China’s single dish telescope.
The second advantage is that the glare problem isn’t as bad. Finding a planet with visible light is hard because the planet is very dim, next to a much brighter star. For radio signals, the Earth may actually outshine the sun by 1,000 times, on certain narrow frequencies. On other frequencies, or during solar flares, the star might still be brighter. But, even if the Earth were dimmer, you could still distinguish the Earth and sun if you used two radio dishes, separated by hundreds or thousands of miles.
The third advantage is that the radio signals could show complex patterns that wouldn’t be found in nature, proving the existence of intelligent life.
So, could any of these radio telescopes detect another Earth?
We need to consider what kinds of radio emissions the Earth gives off, and that’s changed a lot, over time. 200 years ago, there were no radio emissions. 100 years ago there were some radio stations. 50 years ago, the strongest signals were TV stations and military radar. Today, TV transmissions have gotten weaker because we’ve switched to internet and cable TV. But we’ve also added a large network of mobile phone users.
A 2023 paper summed the cell phone and cell tower transmissions across the entire planet to calculate their total radio wave emissions. Most of the radio wave leakage comes from the towers, which are directional antennas that mostly transmit horizontally:
An extraterrestrial observer would see those signals change as the Earth rotates and different continents end up on the horizon:
The exact amount of power transmitted depends on which direction you’re viewing from. The group calculated it for 3 different nearby stars. For one star they concluded:
Our preliminary results demonstrate that for Barnard star, the peak power from mobile towers with LTE technology is in the order of ∼3.2 GW, out-powering the other two mobile technologies
Then they calculated how much power a medium sized radio telescope could detect, at a distance of 10 light-years. They concluded that telescope could only detect a signal with a strength greater than 14,000 GW (gigawatts).
That means you could not detect the Earth’s cell towers with a moderately sized radio telescope. You’d need a signal 10,000 times as strong as our collective cell tower emissions before you could detect it.
The paper went on to calculate it for the single kilometer array, our largest planned radio telescope (still in construction). That should be able to detect a signal with a power of 300 GW. But our cell tower emissions are still 100 times weaker than that.
Back in 1971, NASA and SETI proposed building an even larger array called Project Cyclops, with one thousand 100 meter dishes spread across the planet. That would add up to about 8 times as much collecting area as the square kilometer array. It would almost get you to detecting the Earth at 10 light years away. Maybe for a civilization emitting more radiation than Earth, it would get you all the way there (I’m not sure if the aliens use 4G or 5G or something else).
Unfortunately, congress refused to fund Project Cyclops.
An older paper, from 1978, considered the kinds of transmissions at that time, and the power that they produce. They summarize it with this table:
In short, the FM radio and TV broadcasting networks of 1978 radiated less power than today’s cell phone networks do. But military radar looking for missiles (BMEWS = ballistic missile early warning system) emits pulses stronger than anything else, up to about 200 GW.
Those radar signals would be within the detection range of the single kilometer array.
The 1978 paper concluded that TV stations could be detected by the Arecibo telescope up to 2 light years away, and BMEWS radar up to 20 light years away.
Had the Cyclops project been built, it would have let us detect TV stations up to 25 light years or BMEWS radar up to 250 light years.
If you could detect TV stations, you’d be able to determine the orbit of the Earth, the size and rotation rate of the Earth, and make a map of where the stations were on the Earth.
Just detecting the signal wouldn’t be enough to watch alien TV. For that, your dish would need to be 100 times larger still. The paper’s authors concluded:
Depending on one’s opinion of the information content of television broadcasting, this calculation can be taken either as discouraging or reassuring.
SETI isn’t eavesdropping, they’re looking for beacons
It’s not really in our technical abilities to eavesdrop on another planet, to detect their cell phones or TV signals. It might be possible, if we invested more money in science and built bigger telescopes.
SETI has primarily been looking for something else. They think that advanced civilizations might send out signals to advertise their existence. They’ve hypothesized that the best frequency range to do this on would be between 1,000 and 3,000 MHz. Most of SETI’s observing time has been spent looking at this frequency range.
Some cell phone networks fall in this frequency range, but others do not. TV transmissions and military radar are not in this range, those are in the 40–800 MHz range.
Many stars have been scanned thoroughly by SETI. Congress cut SETI’s funding in the 90’s, but one project was revived by private funding. Project Phoenix went on to look at 800 star systems, all within about 200 light-years away. For each star, they looked at 2 billion narrow channels, between 1,200 and 3,000 MHz.
Project Phoenix didn’t find anything. That doesn’t mean that no one’s out there, though. It just means that less than 1 in 800 stars have planets that are sending us strong radio transmissions in that range. To be picked up, I think these would either have to be focused radio transmissions directed at the Earth or extremely strong unfocused transmissions.
Another SETI project also looked for alien transmissions. If aliens were transmitting to the Earth, using an Arecibo size dish, between 1,100 and 1,900 MHz, SETI could find it within 415 light years. If the aliens were using something 1,000 times stronger, SETI could find that 13,000 light-years away (about 15% the size of the galaxy). They haven’t found evidence of either.
None of SETI’s work proves that aliens don’t exist. It just proves there aren’t any nearby aliens actively trying to send us messages, in the frequency range that would be the most efficient.
We’re not busy sending the same kind of messages to other stars. Advertising our presence to aliens might not be the best idea.
Searching for lasers
Aliens would not have to use radio waves to communicate. They could use some other kind of light, like laser beams.
In theory, you could use a 10 meter wide mirror and a high powered infrared laser, point it at another star, and make a signal that would appear 1,000 times brighter than the sun. We could use that approach to announce our presence to any other solar systems within 100 light years.
This wouldn’t be an accidental transmission, we’d need to intentionally point these lasers at other stars, and even adjust the angles slightly to account for how much those stars will move in the years between now and when the laser beam will arrive. But the technology to do this isn’t complicated — we could already send such laser beams, and more advanced civilizations could as well.
We’ve tried using telescopes to look for those kinds of lasers. Scientists have looked at 5,600 star systems, but didn’t find any such lasers. Those scientists calculated there were about 2,000 habitable planets in that search. They concluded that less than 1 in 2,000 planets in the galaxy are both civilized and advertising their presence with lasers.
You could take a guess as to how many habitable planets are civilized — maybe that’s only about 1%. In that case, they only looked at 20 civilizations, and less than 1 in 20 civilizations uses lasers to try to contact other planets.
Again, that doesn’t prove that those stars aren’t civilized—it’s not clear whether advertising your presence to aliens is a good idea, either with lasers or radio waves.
Searching for dyson spheres
We’re currently trying to replace fossil fuels with solar power. To have a very advanced civilization, we might end up covering large parts of the Earth’s surface with solar panels.
A really advanced civilization might go beyond that and build solar panels in space, or even build a dyson sphere, a shell of solar panels around their star:
If some advanced civilization did build that, we’d be able to detect it: the sphere would look like a large star giving off infrared radiation.
Astronomers have searched for dyson spheres and haven’t found any. But that only gives an upper bound, based on how well they looked. One study concluded that less than 1 in 10,000 stars in our galaxy could have dyson spheres built around them.
We don’t really know who’s out there
Our current technology is still too limited to answer the question.
We could not find or photograph another planet just like Earth.
We could not detect their cell phone or TV signals. We could perhaps detect their military radar signals.
SETI has concluded that no nearby planets are trying to contact us with radio transmissions in the 1,000 to 3,000 MHz range.
We’ve found no evidence of infrared lasers pointed at the Earth. Less than 1 in 2,000 planets are advertising themselves that way.
Less than 1 in 10,000 stars have dyson spheres built around them.
Those kind of numbers give us some limits on how many highly advanced civilizations there are. But we can’t rule out simpler civilizations, civilized but radio-quiet planets, or planets with life but no civilization.
There is a bit of a paradox, here. In my article on the drake equation, I showed that if there are only a few civilized planets in the galaxy, then those are likely very far apart:
If there are a lot of civilized planets out there, that also implies that each civilization tends to last a long time.
SETI’s work does provide some limits — it would be somewhat strange if there are ten million civilized planets in our galaxy, each of them has been civilized for millions of years, but none of them have tried communicating in ways that are easy to receive.
On the other hand, if there aren’t ten million civilized planets, then the closest one is probably further away, and we need to keep building larger telescopes to have a chance of detecting a fainter signal.
