One day, I am randomly teleported to somewhere else where I meet some aliens. These aliens are very technologically advanced and after figuring out communication, we establish a rapport. The aliens, being the benevolent type, want to help me get home, and, as luck would have it, they have a portal device which can spontaneously open a stable wormhole across any distance.

Unfortunately, the aliens don't know where Earth is, and I'm reasonably certain that I'm still somewhere in the milky way--I used to have a photo of the galaxy as a desktop wallpaper and I recognized it when the aliens showed me.

The parameters:

  • The Aliens are isolationist and almost all their data about space was gathered with telescopes. As such, data may be hundreds of thousands of light years out of date.
  • The Aliens have mapped every single star and most larger planets (in the Milky Way) telescopically. More specifically, if you knew which entry was Earth's on their database, it would show detailed information about the Sun along with stating that there are two gas giants (with mass estimates) and between 4-8 other smaller planets.
  • The aliens haven't received any radio transmissions from Earth (I am apparently over 200 ly away)
  • The aliens are willing to teleport out probes to search for radio signals or take a closer look at a some systems but their generosity is finite--they won't saturate the entire galaxy to find Earth.
  • My knowledge about the solar system is at an above average highschooler's level. I can list and describe planets in the Solar System, I can identify or approximately draw several constellations, and I know various random trivia such as the fact that the Moon's apparent size is the same as the Suns'. I don't know any exact numbers like planetary masses or diameters and I don't know anything about the specific type of light the sun gives off, although provided with examples, I would be able to tell what light temperature "feels right".


What search strategy should the aliens employ to find Earth as fast as possible and with the least amount of probes?

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – L.Dutch
    Commented Jul 9, 2020 at 14:13
  • $\begingroup$ 1. If you think there are two gas giants in our system, you will never get home: there are four. 2. Some trivia you probably know if you are interested in astronomy: a. the mass of the Sun is 1000 × the mass of Jupiter b. the apparent size of the Sun is 0.5°, or about the width of a pinkie nail as seen at arm's length. $\endgroup$ Commented Jul 10, 2020 at 11:03
  • $\begingroup$ @EdgarBonet the masses of Neptune and Uranus are small compared to Jupiter or Saturn, from far away it would be significantly harder to detect them. $\endgroup$
    – Dragongeek
    Commented Jul 10, 2020 at 11:07
  • $\begingroup$ If they can see “4–8 other smaller planets”, they can see Uranus, Neptune and Earth. If they can't see Uranus and Neptune, they certainly don't have Earth in their catalog. $\endgroup$ Commented Jul 10, 2020 at 11:11
  • $\begingroup$ I wouldn't trust your aliens. You happened to teleport there and they have a teleporter? Sounds like they're taking the piss. $\endgroup$
    – Muuski
    Commented Jul 10, 2020 at 22:29

13 Answers 13


Instead of searching for Earth or Sun directly, you could scan for notorious star systems close to Earth. As an atronomically interested person, you probably know some bits about some of those, due to their interest for future human space exploration.

The Alpha Centauri system is a triple star system, so through the properties of all the stars (as well as you can remember them) and their constellation with each other you have a lot of properties you can search the database for. Similarly, Sirius is a binary star and a very bright system in the local galactic area and very close to both Sun and Alpha Centauri. You can search for two systems with these properties in close proximity, and then scan the nearby systems for a single star system which matches Sun in colour, number of planets and gas giants.

I'm not an expert on astronomy, so there might be even better "anchor" systems nearby that I'm not aware of.

  • 3
    $\begingroup$ Alpha Centauri and Sirius are both quite common so they won't actually be that helpful of landmarks. Being a triple star system, Alpha Centauri is even more common than a stand alone yellow star like Sol. While Sirius is the brightest star in our sky, it is only a class A star. Other than perhaps Class-O stars, it's really hard for any stellar body to really be rare enough to make for much of a landmark on the scale of entire galaxies, and we don't have any class Os within 1000 ly of us. $\endgroup$
    – Nosajimiki
    Commented Jul 7, 2020 at 22:02
  • 1
    $\begingroup$ @Nosajimiki while there are 10,000(ish) G-type stars like Earth, how many of those will have as their closest neighbour a triple star system between 4 and 4.5 light years away? Rule out all of the G-type stars where the closest neighbour is something else AND all those where the closest neighbour is not within the range bracket and the number should be getting manageable. (Based on your comment, we can also exclude any within 1000 ly of a class O to manage the list further.) $\endgroup$ Commented Jul 8, 2020 at 1:29
  • 2
    $\begingroup$ @KerrAvon2055 There are estimated to be 40 to 80 billion G-type stars. $\endgroup$
    – Schwern
    Commented Jul 8, 2020 at 6:24
  • 1
    $\begingroup$ @KerrAvon2055 Binary stars are extremely common, they may be more common than single stars. Triples are also very common. Of the roughly 50 closest star systems, four are triples: Alpha Centauri, Ez Aquarii, Epsilon Indi, GJ 1245. $\endgroup$
    – Schwern
    Commented Jul 8, 2020 at 6:56
  • 1
    $\begingroup$ @KerrAvon2055 knowing the range to the closest O class star is not something even an "above average" student would know off the top of thier head. $\endgroup$
    – Nosajimiki
    Commented Jul 8, 2020 at 14:52

1. Draw the Milky Way

You mention that you know what the Milky Way looks like. This is good, because it won't actually look the same way from different points in the galaxy - from closer to the centre, it would take up more of the sky, from further above the plane of the spiral, it would show a different shape.

So start your search by simulating its general shape in its sky at different points in the galaxy and picking the ones that look the most familiar. This will rule out maybe 90% of the galaxy.

2. Make a long-list

Now search the alien database for star systems within the possible zones you've identified, and list out all the star systems with the right number of planets.

Filter based on any other facts you can remember - for instance, the sun isn't a binary star, which might rule out some systems.

3. Send your first few probes

Send out a probe to a random entry on the long-list, and check for Earth's radio transmissions. Assuming you don't get lucky first time, take some good pictures of the sky as seen locally.

Repeat this a few times, but make sure the locations you pick are as spread out as possible across the galaxy.

4. Update the star charts

You now have a bunch of star maps from different perspectives in the galaxy. Confusingly, these will be subject not just to three-dimensional parallax effects, but a kind of "four-dimensional parallax" (there's probably a proper name for it) because they'll be seeing the same features with different light-speed delays.

Nonetheless, you should be able to simulate approximately what the sky would look like from places in between, based on the aliens' detailed data and the perspective correction from the probes.

5. Find a familiar sky

Scroll through the photographed and simulated skies until you find one that has some features you think you recognise. Send a probe to a system near that location. As before, listen for radio transmissions, and if you haven't found Earth yet, take some more photos to check and update your constellation simulator.

As you're now looking in much more focussed areas of the galaxy, you'll hopefully find Earth before the aliens' patience runs out and they put you in a zoo.

  • 4
    $\begingroup$ That 4D parallax thing is an interesting point, and idea. $\endgroup$ Commented Jul 8, 2020 at 0:37
  • 5
    $\begingroup$ Spatio-temporal parallax, perhaps $\endgroup$ Commented Jul 8, 2020 at 0:38

There are believed to be about 100,000,000,000 to 400,000,000,000 stars in our galaxy.

That is a lot of stars to check. If it takes one second to check if a star is the Sun - and it might take a loot more time than that - you can check 60 stars per minute, 3,600 stars per hour, 86,400 stars per day, and about 31,557,600 stars per year. Thus you could check 315,576,000 stars in a decade, about 3,155,760,000 stars in a century, and about 31,557,600,000 stars in a millennium. Thus it would take about three thousand to twelve thousand years, to check every star in the galaxy at the rate of one star per second, which seems much too fast for me to believe.

When I was a child I read Rusty's Space Ship, by Evelyn Shibley Lampman, 1957. Two Earth children, Rusty and Susan, help a lizard like alien find his way back to his home at Eopee. They travel from planet to planet to planet in our solar system, trying to find Eopee. Finally the alien happens to sing in his sleep:

A messenger from Eopee

in Andromeda galaxy

my training is the very best

I passed with honors every test

And they wake him up and tell him what he sang in his sleep. So he takes them back to Earth and tells them he will be able to find his way back to Eopee now.

And a few years later, when I knew more about astronomy, I began to doubt whether the alien ever got back to Eopee. The Andromeda Galaxy is even bigger than the Milky Way Galaxy and has even more billions of star system. Unless the alien clearly remembered the directions back to Eopee, he would be forever trying to find a needle in a haystack searching for Eopee in the vast Andromeda Galaxy.

Go back to the beginning. How much knowledge of astronomy does someone need to tell if they are in the Milky Way Galaxy?

Maybe they will ask the aliens to show them pictures of the nearest galaxies. Maybe they will recognized images of the large and Small Magellanic Clouds and the Andromeda Galaxy M31 and the Triangulum Galaxy M33. And if the images are clearly taken from the same angles as telescopic images of them from Earth, then the aliens must be in the Milky Way Galaxy.

There is a theory that planets suitable for life have to exist in within a broad ring shaped area of the galactic disc of the Milky Way Galaxy, not too close or too far from the center of the galaxy. So once the aliens agree that you come from their galaxy, if they believe that theory then they will assume that Earth is within that ring shaped region of the galactic disc, eliminating tens or even hundreds of billions of stars, though tens or hundreds of billions of stars will be left.

If you can compare Earth spectral types of stars with the alien's system of classification, and if you know that Earth is a spectral type G0V or G2V star, and if you can find what type of stars those are in the classification of the aliens, you can eliminate maybe 90 or 95 percent of stars in the zone. And if you tell the aliens that our Sun is almost certainly a single star that should help the elimination.

If the aliens can deduce how much radiation Earth gets from the Sun from an examination of you, and if they can figure out what type of star the Sun is in their classification and thus its approximate absolute luminosity, they can estimate the length of Earth's year with a rough range in their timescale.

If you tell them that Earth astronomy books claim that the disc of the Milky Way is about 100,000 (Earth) light years in diameter and about 1,000 or 2,000 (Earth) light years thick near Earth, comparing those figures with their figures and with the approximate length of an Earth year and thus of an (Earth) light year compared to their units of measure, that will be helpful.

And doing so will be really helpful if you also remember that Earth is supposed to be about 25,000 to 28,000 (Earth) light years, from the super giant black hole at the center of the Milky Way Galaxy. That will enable the aliens to draw a ring around the super massive black hole in Milky Way Galaxy with a radius of their distance units that is equivalent to 25,000 to 28,000 light years. And then the aliens can draw an inner and an outer circle approximating the degree of uncertainty in those figures.

The ring of possible positions of Earth derived from the distance of Earth from the super giant black hole should be a much narrower ring than the ring deduced by the fact that there is life on Earth, so it should be very helpful to the aliens in narrowing down the field of search.

I note that as seen from Earth, The Andromeda and Triangulum galaxies are more than 90 degrees from the galactic center, while the Large and Small Magellanic Clouds are on the other side of the Galactic center and about 60 degrees from the galactic center. So if you can remember that, it would be a good indication which side of the galaxy Earth is in.

An Astronomical Unit (AU) is the average distance between Earth and the Sun. Astronomers measure the angle to a star at two times, six months apart, when Earth is on opposite sides of the Sun and is 2 AU apart. A distance of one parsec is the distance at which a separation of one AU would have an angle of only one arc second. A parsec is equal to about 3.26 light years or 206,264.806 AU.

Back in the 1830s, using a baseline of only 2 AU for their measurements, astronomers were able to measure the tiny six month changes in the positions of stars in the sky accurately enough to find the approximate parallaxes of three stars, and thus their distances from Earth. 61 Cygni is about 3.4947 parsecs, or 11.398 light years, or 720,833.59 AU from Earth, Alpha Centauri is about 1.340 parsecs, or 4.37 light years, or 276,363.5 AU from Earth, and Vega is about 7.68 parsecs, or 25.04 light years, or 1,584,113.6 AU, from Earth. And astronomers were able to measure their approximate distances with the instruments available in the 1830s.

And today Earth astronomers have much better instruments to measure tiny angles much more precisely and accurately. And some such instruments have been packed aboard the Gaia spacecraft.

Gaia is a space observatory of the European Space Agency (ESA), launched in 2013 and expected to operate until c. 2022. The spacecraft is designed for astrometry: measuring the positions, distances and motions of stars with unprecedented precision.[7][8] The mission aims to construct by far the largest and most precise 3D space catalog ever made, totalling approximately 1 billion astronomical objects, mainly stars, but also planets, comets, asteroids and quasars among others.[9]


If the aliens have the ability to send equivalents of the Gaia spacecraft, or much more advanced versions of it, through wormholes to any place in the galaxy, they can have pairs of those automated observatories observe a star from locations that are many light years or parsecs apart, and thus with baselines, tens, and hundred of thousands, and millions, of times as long as the baselines used by Earth bound astronomers, and they can measure the directions and distances to that star, and thus its position in three dimensional space, much more accurately than Earth bound astronomers. And they can do that for all of the billions of stars in our galaxy. And by measuring the apparent magnitude of a star at a specific distances, they can calculate its absolute magnitude, and thus calculate how bright it will appear from any point in the galaxy.

So the aliens will ask you about prominent constellations as seen from Earth. And with luck you should be able to draw not too bad depictions of a few, such as the Big Dipper and the little Dipper. And the most prominent Constellation, Orion. And by great good luck the distances to the stars in Orion which are brightest and most visible as seen from Earth happen to differ much less than the distances to the brightest stars in most constellations do. Thus Orion will be recognizable at a much greater range of distances than most Earth Constellations are.

And if you happen to remember that, as seen from Earth, Orion is almost opposite to the center of the Galaxy, the aliens should come up with a smart idea. If the aliens select points in the ring around the center of the galaxy where they think Earth should, be at one degree intervals as seen from the center of the galaxy, each point should be only about 440 to 490 light years from the next. And that should be close enough.

Then for each of those 360 points they can generate images of the sky as seen from that point, opposite to the center of the galaxy, with the apparent positions and the apparent magnitudes of the stars calculated. And at least one of those points should show a rather distorted image of Orion. And they they can calculate the appearance of the sky from a number of equally spaced points around that point, and find a point where Orion looks most like it is seen from Earth.

And then it will be helpful if you can remember the constellation of Taurus, which appears to the right of Orion from the northern hemisphere. Many of the brighter stars of Taurus form a sort of V shape and are in the Hyades star cluster about 153 light years from Earth. But a star along one side of the V, on the side toward Orion, is Aldebaran, which about 65 light years from Earth. Another star cluster, the Pleiades, about 450 light years from Earth, appears to be on the side of the Hyades away from Orion as seen from Earth.

So a spot where Taurus is recognizable is likely to be close to Earth.

So if the aliens can produce images of Taurus and Orion you recognize, they can send probes to study many star systems in that area of space to see if they are your solar system.

And I can't help wondering if someone with merely average high school knowledge of astronomy would be able to remember all the details which would help the aliens find Earth for them. I can't help thinking that if by chance they had a pocket book of astronomy on them when taken from Earth to the alien's world that would help the aliens a lot.

My answer to this question


discusses how someone with more astronomical information could find his location in space.

  • 1
    $\begingroup$ Best answer. Triangulation from positions of Magellanic clouds, and narrowing down locations by hunting for almost-but-not-quite-matching constellations is the technique I came up with. $\endgroup$ Commented Jul 8, 2020 at 11:08
  • $\begingroup$ You're overlooking (I think) a big timesaver: there may be 400,000,000,000 stars in the galaxy, but the OP is likely to know that Sol is out near the edge, not in the densely populated center. The OP may also remember that the orbital period of Sol around the Galactic center may be around 240,000,000 years, which may also narrow down the search. $\endgroup$
    – chepner
    Commented Jul 8, 2020 at 18:42

If there are details about planets and star characteristics you could use a process of elimination according to planet / star characteristics.

I, like your protagonist, have a knowledge of the universe at an above average high schoolers level, though less sure about "above". But I can use excel.

Here is how I would do it if the alien database were in excel.

I would sort by planets, looking to isolate the subset of planets with either oxygen nitrogen atmospheres, or liquid water or both. Those are Earth candidates.

I would probably turn up a few hundred million. I would sort those planets according to moons; one and only one. Then, assuming that Earth's moon is unusually large I would exclude planets with a moon in the bottom 50% of size range for those moons. Now there are a million candidate planets.

Then I would sort those systems for the presence of 2 and only 2 gas giants. That would narrow it down to tens of thousands.

I would sort systems by star type, excluding the reds and giant stars. Now that is down to a few hundreds.

If I can sort gas giants by presence of a ring or not that would be helpful because maybe Saturn is special.

Now there are a hundred or so Earth candidates. I would look at them one by one. Hopefully the aliens have a visual catalog that makes sense to me of how the different planets and stars look. I would look for blue seas and white clouds. I might be able to identify Luna or Jupiter.

  • 5
    $\begingroup$ The aliens guaranteed two large gas giants (Jupiter and Saturn?) and were vague enough to arrive at only "4-8 other smaller planets". If their detection is weak enough that they can't even tell small planets apart sometimes, I doubt they would have information on moons, rings, atmospheric composition, water, or the like. $\endgroup$
    – parasoup
    Commented Jul 7, 2020 at 16:11
  • 2
    $\begingroup$ @parasoup: I figured if humble earthlings can deduce info about exoplanet atmosphere from spectral lines, the aliens can do that too. en.wikipedia.org/wiki/…. But agreed: if the alien catalog is super lame it will be hard to use it to find earth. $\endgroup$
    – Willk
    Commented Jul 7, 2020 at 16:29
  • 2
    $\begingroup$ You can likely narrow the field further by looking for stars whose closest neighbour is a Ternary system at 4.3ly (Centauri) $\endgroup$
    – Kyyshak
    Commented Jul 7, 2020 at 18:14
  • 2
    $\begingroup$ The Moon is a good bet. It's not unusually large, but it is unusually large in relation to its planet. Every other moon in the solar system is tiny by comparison to its planet, but our Moon is 25% the Earth's radius. Except Charon/Pluto but that's practically a binary system. However, as we haven't seen an exomoon yet, it might not be unusual. Ternary systems are fairly common, 4 of the closest 50 systems are ternaries. $\endgroup$
    – Schwern
    Commented Jul 8, 2020 at 6:59
  • 9
    $\begingroup$ The technique of trying to identify your star by describing its planets is a bit like asking someone in another continent to help you find your house by describing the flowers in your garden. $\endgroup$ Commented Jul 8, 2020 at 11:15

Frame challenge (especially if this is science-based)

"The Aliens have mapped every single star and most larger planets (in the Milky Way) telescopically."

No they haven't. To do this telescopically simply isn't feasible. Apart from anything else, stars are hidden behind other stars. Also our view towards the center of our galaxy is blocked by the HII clouds - or dust. We can see the hidden parts only at certain wavelengths. https://en.wikipedia.org/wiki/H_II_region

There isn’t a way to simply count the number of stars in the Milky Way individually ... https://asd.gsfc.nasa.gov/blueshift/index.php/2015/07/22/how-many-stars-in-the-milky-way/

If you have any doubt, just watch this https://youtu.be/X-3Oq_82XNA

  • 5
    $\begingroup$ This can be mostly solved by a small modification to the premise: they have wormholes, and can send probes. The current question implies they've never used these for astronomy, but if we modify that to rarely, they can map a lot more sky just by sending probes to strategic points to see the other side of major obstacles. $\endgroup$
    – IMSoP
    Commented Jul 7, 2020 at 16:35
  • $\begingroup$ @IMSoP - I haven't done the maths (yet) but even using wormholes it's going to take longer to explore the galaxy than the time these aliens have been in existence - especially if they had to evolve from the basics and then develop their amazing technology. Nothing they see telescopically will tie up with anything they can see through the wormhole. Their database will be huge and the computer time to work out where everything is in relation to everything else is ridiculous. Unless we bring magic into this science-based story, it just doesn't stand up. $\endgroup$ Commented Jul 7, 2020 at 16:43
  • $\begingroup$ H II regions are ionized gas around hot, massive stars; I assume you're thinking of H I regions, molecular clouds and other neutral bits of the interstellar medium. $\endgroup$
    – HDE 226868
    Commented Jul 7, 2020 at 16:58
  • $\begingroup$ @HDE - I bow to your superior knowledge. I think you'll agree though that whichever way we look at it, there's far too much sh*t in the way to make a complete telescopic map. $\endgroup$ Commented Jul 7, 2020 at 17:40
  • 1
    $\begingroup$ @chaslyfromUK I know, that were the same numbers I used to arrive at the around 100k telescopes. Completely feasible for a civilization on that tech level. $\endgroup$
    – mlk
    Commented Jul 8, 2020 at 11:08

How well you can reliably narrow it down

There are an estimated 100-400 billion stars in the Milky Way.

Identifying our Star Type:

Sol is a G-Type Main Sequence Star, or as most people would simply call it: a yellow star. Only about 1/10 stars are yellow so this should narrow your search down to about 10-40 billion possible stars.

Identifying our Galactic position:

This could potentially be the best tool in your tool kit depending on how exact your traveller's memory is and how good his wall paper was. If your traveller only remembers about how far from the core of the galaxy Sol is, then this could get it down to about 1-20 billion possible stars depending on how confidently he can remember this detail, but if your traveler really knows his stuff (which he may or may not based on your description), then he would be able to identify where Orion's arm splits off into Orion's spur, and then be able to show that Sol is somewhere in this area. This should be enough to get within about 5000 ly of Earth. Since there are only about 600 million stars in this area, this would reducing the number of possible star systems to about 60 million.

How the inaccuracy of the alien star charts would foil you

Even 60 million is still too many to count without narrowing it down more, and this is where you run into huge risks of failure. Filtering large lists is all about throwing away results without ever actually looking at them. So, if you apply a single filter to your list of stars that drops Sol, you will never find it; so, you have to be very careful about utilizing the following methods:

Identifying our neighbors:

Most people know that Alpha Centauri is about 4 light years away and is part of a trinary star cluster. This is where you start running into issues. We humans consider Alpha Centauri to be a trinary star system but what happens if the aliens have a different definition of a star system. Perhaps they don't consider class M5 dwarves like Proxima Centauri to be stars at all or perhaps it orbits at too great of a distance to meet thier definition of trinary star system. So, in thier records, Alpha Centauri may be a binary star orbited by what they still consider to be in the range of "thermal emitting gas giants" or something that they have a unique word for for things between "stars" and "gas giants". Or just a binary star with a really close neighbor star. Or, your travel does a bad job describing light years; so, the aliens are actually looking at the wrong distances away. There are all sorts of ways aliens might classify things differently than we do that could cause major issues. This means that if you filter only for G-Type Main Sequence Stars that are about 4ly from trinary star systems, there are multiple possible reasons that your alien database will throw out Sol as a possible match.

Even if you get lucky and don't throw it out, G-Type main sequence stars and trinary star systems are both so common, that you will still have thousands upon thousands of stars left to go through needing yet more filters.

Identifying our planets:

...if you knew which entry was Earth's on their database, it would show detailed information about the Sun along with stating that there are two gas giants (with mass estimates) and between 4-8 other smaller planets.

The number of planets becomes an even bigger issue. Your highschool student does not know how many planets the aliens know about in the Sol system. He also does not know how to accurately describe thier sizes; so, when you query a database containing somewhere between what is now a number somewhere between tens of thousands and 40 billion records (depending on what you've already done to narrow it down) being close is still not good enough. Will the aliens be looking for stars with 8 planets? or is it 9? or maybe 15? See we humans live here and even we can not all agree on how many planets are in our own solar system. Either way, the aliens will be so busy looking at stars with 8-15 planets in thier database that they will never get to some poorly documented star with only 6 known planets in thier database. Because thier records are not perfect, there is a good chance you will filter out our solar system and never find it if you try to look by planets.

Photographic Fidelity:

Let's say the alien database is really good, and your traveler does a great job of explaining things getting you to the point where you can actually narrow down the list to a manageable set that actually includes Sol. What do you do? Well, the most efficient thing would seem to be to look at the photos they have of our solar system. This of course raises the question: could you even recognize it or would you throw it out will all the other near matches? When it comes to taking a picture, there are all sorts of factors that can make the same familiar thing look different. When photographing things in space, how you expose your image based on that tinie-tiny dot of light in the sky makes a huge difference. Below are 3 images taken of Saturn by NASA. If the aliens sat you down to look at a lineup of a few thousand Sol like systems, and they randomly showed you one of the 3 photos, would you know for sure that you are looking at Saturn or just one more of the many many ringed gas giants we have in our galaxy?

enter image description here

How the inaccuracy of human memory would foil you

Identifying our constellations:

If your traveler has an actual picture of our night sky saved on his phone, then it would be relatively easy (with enough computational power) to line up our constellations with our exact position, but these kind of things have to be very precise. If you are just trying to draw constellations from memory, this will more likely than not lead you on yet another wild goose chase. What you draw as the big dipper will likely more closely resemble many many constellations as seen from other stars; so, if your sketch is not accurate enough, you might end up filtering Sol out of your candidate list in trying to narrow it down.

Then there is the bigger problem of actually trying to be too specific. Our constellations are often idealized versions of actual stars where parts of the constellation that "make up the image" are not actually the brightest stars in that part of the sky. So if you don't draw out orion's arm and bow, your image will likely be to vague to find a match, but if you do draw them out, you are in even worse shape because all the missing stars from your sketch will nullify Earth as an option at any threshold of apparent magnitude.

enter image description here

The devil is in the details

This leads to the final problem of human error. What happens if your student remembers anything wrong? He is going off of memory after all, so if he remembers just one detail wrong like maybe he thinks Alpha Centauri is 3.4 ly instead of 4.3ly or maybe he does not remember Mars having moons or something like that. On a test like this, anything less than a 100% is a failing grade.

In summary, it is very unlikely that a traveler is going home without either a much better than high-school level understanding of galaxy or some kind of citable documentation.

  • $\begingroup$ Proxima Centauri is a class M5 star -- very much not a deuterium-burning brown dwarf. $\endgroup$
    – Mark
    Commented Jul 8, 2020 at 3:00
  • $\begingroup$ One presumes that if they have mapped the position and velocity of every single star in the galaxy, they would have them on computer. That presupposes they have the computing power to model what the sky looks like for every star using straightforward image processing techniques. Hence you should be able to draw some constellations and get the software (come on, it has to be at least minimally sophisticated) to identify places where the constellations look like that. $\endgroup$ Commented Jul 8, 2020 at 11:02
  • $\begingroup$ @Mark I'm not sure what article I glanced at that called it a brown dwarf, but M5 stars are at the cusp of red dwarves and brown dwarves; so, they do not get classified the same according to every scale, but I agree my answer was unnecessarily confusing so I just reworded it to "class M5 dwarves". That said, this kind of ambiguity is my exact point. Modern scientists typically use Hydrogen burning as the base line, but what if they use convective patterns or whether or not it can eventually nova? My point that it might not be what they call a star remains. $\endgroup$
    – Nosajimiki
    Commented Jul 8, 2020 at 13:50
  • 2
    $\begingroup$ @PrimeMover If the traveler had his actual high school textbook with accurate photos of our night sky, I would agree, but without some way to confer exact shapes and aspect ratios of the constellations to the aliens, you can't really draw any conclusions based on a rough sketch. If you were to just sketch something like the big dipper, chances are the aliens could point out something similar in thier own night sky. $\endgroup$
    – Nosajimiki
    Commented Jul 8, 2020 at 14:01
  • $\begingroup$ @Nosajimiki Doubtful, very doubtful. The mathematics of probability is against it. And even then, if there were a similar pattern to Ursa Major (I refuse to use nursery terms), there's not going to be another pattern similar to Orion. Oh, and if your OP is so clueless to not know his constellations, he deserves to stay lost forever, I don't want to share my planet with someone so intellectually worthless. $\endgroup$ Commented Jul 8, 2020 at 14:06

I can list and describe planets in the Solar System, I can identify or approximately draw several constellations,

This is what can bring you home, if possible at all. It would be better if, together with drawing some constellation, you could also indicate their relative position in the sky.

If they have a catalogue of all the stars in the galaxy, they should be rather easily be able to build a 3D representation of the same galaxy. From there to calculating the appearance of the sky from any point of view would be doable (it's something we can pull out with our tech level).

Since you know and can draw several constellations, you can either scroll the list to search for star systems similar to the Sun and check what the sky looks like from their point of view, or scan the point view searching for places where the sky shows the features you can describe and then search for Sun like systems in the area. In both ways it would be probably better to have an expert system do the job for you, I suspect there will be a large number of stars to check.

Once you have the list of candidates, you can try to weed it out further and hope it's not longer than the list of probes they are willing to send.

  • 3
    $\begingroup$ I don't think this would work; you could be anywhere in the Milky Way, and it's been stated that these aliens' starmaps might be out of date since they rely on old light. If they have really good proper motion data, then you might be able to scroll through time to see when the constellations line up with how we know them now, but that would lengthen your search time considerably. $\endgroup$
    – parasoup
    Commented Jul 7, 2020 at 15:53
  • $\begingroup$ A change in POV means a change in time. You'd need to know the masses, positions, and velocity vectors of every nearby massive object to project their movement at different times. Any missing data or inaccuracies will increasingly throw off the calculations. Maybe the aliens have that data, but we do not. We only have estimates for a small fraction of the Milky Way. $\endgroup$
    – Schwern
    Commented Jul 8, 2020 at 7:05
  • $\begingroup$ Most constellations consist of stars that are close rather than necessarily being unusually bright. However, some constellations (most notably Andromeda) include other galaxies within them, and of course our own galactic centre ("Sagittarius A*"). If you happened to remember the positions of a few nearby galaxies and the distance to the galactic centre it could be used to approximately triangulate our position within the galaxy (within a few thousand light years), reducing the initial search space. $\endgroup$
    – Steve
    Commented Jul 8, 2020 at 9:11

Expanding further on my comment to another answer...

You remember part of the lyrics of "the galaxy song" from Monty Python's meaning of life (or otherwise have some reason to remember similar statistics):

Our galaxy itself contains a hundred billion stars; It's a hundred thousand light-years side to side; It bulges in the middle sixteen thousand light-years thick, But out by us it's just three thousand light-years wide. We're thirty thousand light-years from Galactic Central Point, We go 'round every two hundred million years; And our galaxy itself is one of millions of billions In this amazing and expanding universe.

This gives the approximate distance from earth to the centre of the galaxy as 30,000 light years, and also calibrates what a light year is for the alien's benefit as 1/100,000 of the distance across the galaxy.

So far this only gives us the (very approximate!) distance from the centre of the galaxy... but you also remember that the centre of the galaxy from our point of view is Sagittarius A*, and of course that the nearest galaxy to us is the Andromeda galaxy in the Andromeda constellation.

The relative position of the galactic centre to the Andromeda galaxy would greatly narrow down the search area. It would be even more useful if you can identify and remember the correct relative location in the sky from Earth's perspective of some other nearby galaxies to improve the triangulation and reduce the search area. In particular a comment from someone more familiar with astronomy than I suggests the Magellanic Clouds - being different distances away, they would only appear at the right apparent spacing from someone in a similar region of our galaxy.

Also, supernovae that you know about (where in the sky they appeared as viewed from earth, and when in earth's history they were recorded) could be useful as further triangulation points if they're ones the aliens also know about. If the aliens know exactly where and when in the galaxy the supernova actually occurred, then the number of years after the event that it was observed from earth would give a distance... but for many of these it could be uncertain WHICH supernova was the one we saw given that there are potentially thousands of [light-]years of uncertainty about the position we were observing from, and large numbers of supernova events that may be observable over such a time span. That could even make an interesting blind alley for the story - the aliens correlate the only supernova you can remember with one that hasn't been observed from earth yet.

The exact shapes of constellations would seem less useful initially, but once you've got an approximate location, the aliens' database could perhaps come in useful to reconstruct what an observer at various places in the search area might see (unlike the questioner who assumed planets would be known about, I'm assuming only the position and brightness of each star is known). This can also be correctly aligned (using the known positions of the galactic centre and andromeda galaxy) so that you know which constellation you should be looking for in which part of the sky as reconstructed from the star charts.

I could imagine it working something like this:

The aliens put together a database of all known stars with their brightnesses and exact positions and velocities to create a simulation of what the sky would look like from any specified position as of the time you left your home system (as other answers, they may need to send some probes to locations not to far from the target area to update their database). They set you up with a few example view points. Every time the stars look totally different. Surely this is hopeless - you won't ever recognise anything. So many images have gone back and forth you can barely remember what you're looking for.

The screen is focussed on the view towards andromeda - one of the few common waypoints you could identify with certainty. The first try with Sagittarius was hopeless, as there are even more stars in that direction and you can make almost any shape you want (besides you can't remember the shape of that one so well).

After flicking between hundreds of random search points within the target area, suddenly "the projected view to Andromeda galaxy from this place kind-of looks a little bit like Andromeda, but really distorted and with stuff missing", then the alien lets you jiggle the stick a little, and a star approaches from the right side of the screen. You try to use the star to complete the picture, but it's still badly distorted. Looking at other parts of the simulated view from this place, nothing is recognisable. Disheartened, you think it must just be a co-incidence. The alien moves to a full 360 degree projection room and moves the simulated view point back and forth along the direct line to Andromeda. The star you picked fades almost completely out of view, and whichever way the alien moves the view, it gets worse. "It's no use - it must have been a co-incidence, the rest of the sky is totally wrong". As you turn to say this, the stars whizz past to the side. The galaxy never seemed so big before, even some of the smallest movements completely change what you see. Suddenly something familiar is there, and then gone again. "Go back a bit" you scream. Suddenly there it is - not quite right, but recognisable - the most recognisable constellation of the northern hemisphere, Ursa Major. "fix on that, and move really slowly". The distortions get worse. "No, the other way". The sky starts to come into alignment. Looking back towards Andromeda, the star you'd picked is barely visible and well out of position now, but another has moved over and is now in almost the right place.

The alien reveals how lucky we got - in the whole search we'd only panned the viewpoint by less than a hundred light years. The triangulation had several thousand light years of uncertainty. Never before had you appreciated just how big the universe was.

You look again at the chart. Something is very wrong. "I don't remember a huge bright star there." The alien looks up the statistics for that star. "The simulation says this is within 1 of your light years of the simulated position. That star should have been REALLY obvious". Suddenly you realise the problem... "can you delete that star from the simulation and set the view point to exactly the position of the deleted star?". The last few distortions are removed - you are finally "home".

"I'm glad you found it, but you must have remembered something wrong" says the alien. That wasn't even supposed to be in the search area. The star you've selected only seems to be less than 26,000 of your "light years" from the galaxy centre.

(put another way, the galaxy is huge, so even with this kind of triangulation to get something approximating the right search area, you're going to need to have some way to inject some lucky fluke into the story).

  • $\begingroup$ Better than Andromeda may be the Magellanic clouds -- together with the main galactic mass they provide a specific asymmetry with which a seasoned astronavigator can always get his or her bearings. $\endgroup$ Commented Jul 8, 2020 at 10:55
  • $\begingroup$ @PrimeMover updated to mention this, thanks for the suggestion. $\endgroup$
    – Steve
    Commented Jul 8, 2020 at 11:06
  • $\begingroup$ Re-reading your colossal screed, I just noticed your comment about supernovae. Not only must the Crab Nebula be really distinctive, but also its age is fairly well-known. Also, the seriously weird anomaly that is Eta Carinae may also be on their radar, so to speak. This question never stops giving. Someone's going to have to write the story. $\endgroup$ Commented Jul 8, 2020 at 11:12
  • $\begingroup$ @PrimeMover to be fair - the comment about supernovae wasn't there the first time you read it. I'd guess the key to a good story would be combining hard astronomical facts with stuff that someone who hasn't memorised a tonne of astronomical data would actually be able to remember - to some degree. $\endgroup$
    – Steve
    Commented Jul 8, 2020 at 12:26
  • $\begingroup$ @PrimeMover the flip side of "supernova that's not been observed from earth yet" I mentioned could be that the supernova that caused the crab nebula hasn't been observed from the alien home world yet, making a highly memorable event for earthlings that doesn't even feature on the alien database... which would explain why they tried to correlate it with a supernova known to them but not to us. $\endgroup$
    – Steve
    Commented Jul 8, 2020 at 12:33

Condition A
You know our star is a yellow medium sized star with two gas giants (we don't know if the aliens managed to detect Uranus and Neptune). It's closest neighboring system is trinary system. Alpha Centauri-A is similar to our Sun while Alpha Centauri-B orbits it closely is of the same type but smaller and less bright. Alpha Centauri-C is a red dwarf that orbits them at a large distance. At the moment the confirmed planets in the system may not have been detected by the aliens (no gas giants).
So the question is: how many combinations like this are you going to find in the galaxy? At the moment we can't answer this but considering G stars like ours are not that common the answer may be not that many.
Still it's a 'not that many' of more than 100 billion stars. We need to add another condition to filter them more.
Of course if the aliens' data are extremely ancient Alpha Centaury may not be our closest neighbor yet.

Condition B You know that our system is not in the core of the galaxy where star density is highest.

Bottom line: The sun is about 1/3 the distance from the center of the Milky Way galaxy to its outer edges. It’s located in a smaller spiral arm, between two large arms, called the Orion Arm. earthsky.org

Now, how many systems satisfy both conditions? Hopefully not many. Maybe just one.

I assume the aliens have a system to do queries easily on their star map. Also that the above two conditions are known to your character. Otherwise he went to the wrong high school.

  • $\begingroup$ That Alpha Centauri is our neighbour and more important that it is a trinary, I think those bits are not high school. There is an astonishing number of people who just know the barebones because they can't muster any interest. $\endgroup$
    – Anderas
    Commented Jul 7, 2020 at 18:13
  • 1
    $\begingroup$ @Andreas possibly. I remember me talking about Alpha Centaury with a friend. He actually explained to me that Proxima Centaury (the red dwarft) is the closest to us. We were in elementary school. 9, 10 years old. $\endgroup$ Commented Jul 7, 2020 at 18:57
  • 1
    $\begingroup$ Two things: first, the Sun is a class "G" star, not a class "M". And second, class "G" stars are only uncommon in comparison to the ubiquitous red dwarf stars -- they only make up about 7.5% of the galaxy. $\endgroup$
    – Mark
    Commented Jul 8, 2020 at 2:57
  • $\begingroup$ Sun class corrected thank you. I get the point of stars like ours being relatively common. That is why condition B is there too. How many systems in the galaxy will satisfy both conditions? hopefully few enough that you can focus on them for your research $\endgroup$ Commented Jul 8, 2020 at 8:15
  • $\begingroup$ The Sun is not about 1/3 of the distance from the center of the Galaxy to the outer rim, it is about halfway. If the character can't remember Earth's position better than some of the answers he's in big trouble. $\endgroup$ Commented Jul 8, 2020 at 17:51

Very interesting question, I hope you process it into a 7-tome opus about a searching for Earth. Something with "finding" or "founding" in name :-).

I will assume that you are in our galaxy, because there are really too many galaxies in our visible universe and many of them will perfectly match our rather limited knowledge of Milky Way's actual shape and your vague recollection of it. It's easy to explain away with a limited portal range anyway.

What other answers mostly forgot is the need to define to the aliens our units of measure. How long is the light year? How long is the year? And so on.

Units based on human experience, like the weight of a piece of metal in Paris or a length of Earth's meridian, were in the last decades mostly replaced with objective and independent units, but I guess you do not remember their definitions.

But never mind, you still have YOU as a measure! You probably know your height, weight and temperature with a good-enough precision to define kilogram (would it be needed), meter a kelvin. You have to also estimate your pulse as 80/min, which will give you the second and the light year.

(It would also work if you knew only US traditional units, but it would be better to remain lost than admit to the aliens that you still use them ;-))

The Sun is orbiting our galaxy's center at approximately 25000 ly, close to one side of the thin disc. You might also remember relative positions of galactic center, Magellanic clouds and the Andromeda galaxy. That would roughly show Sun's position.

You could also remember approximate shape of the orbit of the S-2 star in the galactic center, which again very roughly reveal our galactic quarter.

Observations of supernova 1987A and Kepler's supernova of 1604 would rather precisely show our position (on a circle), but as others pointed out, the aliens' world can be shifted thousands of years relative to these events, so there's no way to find out if you and them talk about the same supernovae.

Sun is a single star, yellow dwarf with a surface temperature about 5000K, 5M years old. Its closest neighbor is a ternary star at approx. 4 ly.

At the end, when you search row after row in the aliens' star system spreadsheet for the one matching above criteria, with the right amount of planets, look for the one with a note "Mostly harmless".


they'd have their work cut out for them. because they'd have over 10,000 stars to scan. and out of those, you'd have 200 yellow dwarfs (G type stars) to look over. as for constellations, forget it, and for good reason: the night sky will be so different 200 light years out, that anything you draw would be useless in that they'd looked totally different from their own cataloged constellations.

And there's a another factor: Year length of different planets. the year length and day and night cycles of where you'll be at will be different enough that you'd be hard pressed to try do calculations to determine the year length of different planets they find.

but there is hope, just find catalog of a yellow dwarf who's system data matches the sol system, and then have them search there again.

  • $\begingroup$ I don't understand what you mean about year length. Are you saying there would be no way to determine a baseline for converting their time units to ours? I think there's a question about that somewhere... $\endgroup$
    – IMSoP
    Commented Jul 7, 2020 at 16:32
  • $\begingroup$ i am just saying it'll be difficult at best. Also real world examples of two varying measurement systems that exist are the Imperial system and the metric system. plus how humans count is based on the number of digits per hand. and year length is the time a planet takes to orbit its sun. which can very from only a few standard days, to several standard centuries taking all of this into account, i can safely say that it's entirely possible that trying to form any kind of baseline may very well be virtually impossible. $\endgroup$ Commented Jul 7, 2020 at 17:35
  • $\begingroup$ Can you count out seconds? The aliens will know the speed of light; a meter is about one-three-hundred-millionth of a light-second. You'll be accurate to within a few percent, if you can keep time well. $\endgroup$
    – parasoup
    Commented Jul 7, 2020 at 17:55
  • $\begingroup$ that's assuming they have numbers up to 7,8,9, &10. and example is if they have eight digit hands, the number they'd have as a base would go up to 8 before becoming 11. that would mean they'd have a system of mathematics based on denominations of 8 instead of our system based around denomination of 10. if they entire species can only count up to 8, there'll never be such a thing to them is the numbers 9, or 10. $\endgroup$ Commented Jul 7, 2020 at 18:33
  • 2
    $\begingroup$ It would not be hard to teach them how decimal worked. If they have the math for interstellar teleportation, I would be astonished if they did not have a notion of different bases in arithmetic. If aliens started trying to count to me in hexadecimal, I could convert the numbers they used into ones I could understand pretty easily. $\endgroup$
    – parasoup
    Commented Jul 7, 2020 at 18:35

Assuming you actually are in the Milky Way (some answers point out, you might not recognize it)

You should also know that Earth is in an Arm near the edge of the Galaxy. So your potential search area is limited to the edge. Update See comments below and other answers, humans really don't know how big the Milky Way is or how close Earth is to the edge of the Galaxy.

Assuming the Aliens have technology to pick up and isolate weak radio signals....

The Earth has a broadcast influence of about 100 light years. This space is occupied by about 14,000 stars.

If you drop probes with a spacing of about 100 light years, and listen you should be able to pick up radio and T.V. signals, some of that will be in English a language which they can compare, using your speech as an example.

Once you have found the correct 100 Light year sphere, it should be simple enough to narrow it down to find Earth.

  • $\begingroup$ Earth is not near the edge of the galaxy, but about halfway from the central point to the other edge of the galactic disc. Thinking of looking for Earth broadcasts is a good idea. But since the galactic disc is about 100,000 light years in diameter you would need 250,000 probes to cover the whole disc, multiplied by at least 10 times for thickness, & multiplied several times to ensure adequate overlap. And possibly multiplied because the broadcasts might not be recognizable for a full 100 light years. So you better hope the aliens are willing to send out that many probes to help you. $\endgroup$ Commented Jul 8, 2020 at 17:45
  • $\begingroup$ @M.A.Golding The image I remember of our location is similar to this one clarkscience8.weebly.com/solar-system-exploration.html but this image phys.org/news/… suggests we are farther in then you propose. All of which provides support for idea that, a human today is unlikely to recognize their own galaxy from an image. Let alone identify, Earths location in the galaxy if you have the right one. I do still think checking for Earthly broadcasts is going to be your best option. $\endgroup$ Commented Jul 8, 2020 at 18:01
  • $\begingroup$ @JamesJenkins You realise you have to be within 100 light years to be able to detect any of these Earthly broadcasts? That means, on the galactic scale, really really close. And if you're that close, you're home and dry by identifying constellations. $\endgroup$ Commented Jul 8, 2020 at 22:30

If you have a photo of Milky Way on your bedroom wall, you presumably will have gazed at it while lying in bed, dreaming of travel, and undeniably will have a "You are here" label on it somewhere.

It is to be assumed that if it is a genuine photograph from outside Milky Way, it will also have the Magellanic Clouds in the shot as well.

So, if you are able to look at such a photograph (or even 3-d map) with your friendly aliens, you will see exactly where we are situated in that galaxy merely by using those Magellanic clouds as navigational aids. Then it is straightforward to remember where Sol is situated with relation to the specific spiral arm, and that will narrow it down far enough for you to be able to be placed in a similar enough vicinity that in some direction (in front and behind) some of the constellations are vaguely recognisable. Your home lies along that axis.

Both Jack Vance (in his Demon Princes series) and Isaac Asimov (somewhere in his massive oeuvre, can't place exactly where, I'm less intimately familiar with his work) discuss exactly that question. Vance brushes it off as a technique that any vaguely competent space-traveller would have as a tool under his belt.

  • $\begingroup$ No 21st century Earth human has real photo of the Milky Way Galaxy taken from outside. With our current technology it would take hundreds of thousands of years to acquire such a photo. $\endgroup$ Commented Jul 8, 2020 at 17:48
  • $\begingroup$ @M.A.Golding The point being that if he had a "real photo" of the Milky Way, OP is living in a society in which extra-galactic travel is matter-of-fact and de rigueur already. That's the angle I was coming in on. Fact is, OP's scenario is a swiss cheese you could drive a stupid bus through. $\endgroup$ Commented Jul 8, 2020 at 22:17

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .