In my world, humanity reaches their new home among the stars by way of a portal that pops into the solar system one day. This portal instantaneously transports those who enter to a place that is located a jaw-dropping 6 gigalightyears (six billion lightyears) from Earth.

But the astronauts traveling there don't know that. Would it even be possible for them to determine how far they had been transported? And in what direction?

I would imagine that they'd look for known galaxies, or perhaps look at the cosmic microwave background for clues. But those things might appear radically different, considering the astronauts have been transported over a billion years into the past (or is it the future? Wormholes confuse me).

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$ Commented Dec 8, 2017 at 4:54
  • $\begingroup$ Do they already know it's 6 billion light years from Earth? If they know the distance that would solve at least one part of the problem. $\endgroup$
    – Schwern
    Commented May 4, 2018 at 18:26

16 Answers 16



The OP updated in a comment that communication is possible through the portal. My answer assumed that you had to figure out where you were from the other side, using only the information you brought with you.

If you can do observations from both sides of the portal, then there is a way to find out where the other side is, so long as the event horizon of the other side overlaps with Earth. At 6 Gly, the event horizons should overlap. EricTowers' answer provides a way of doing so.

Hard No

6 giga-light years is a very long distance. The fundamental problem with identifying anything if you were mystically transported that distance away, is that the scale of light years also corresponds to a scale of years. That is, traveling such a great distance also causes you to effectively travel through time.

Travel through space is travel through time

For an example of what I am saying, if you were 6 billion light years away and could somehow see the Earth across that distance, you would be seeing the Earth 6 billion years ago. That is, you would see the Earth before the sun even formed, so you wouldn't really see Earth at all, just a cloud of interstellar gas. In fact, you would effectively have traveled to a 'time' when Earth doesn't exist for you; Earth's existence in space and time is outside your event horizon after going through this portal.

The universe is different at a different 'time'

Applying this to the cosmos at large, you will fundamentally not be looking at the same universe that we can observe from Earth. At only a few points in the universe would you be able to observe the things that you can see from Earth at the time you left Earth. Assuming simple Euclidean geometry (note: not a good assumption in this case! but easy to understand), the only things that you can see that you could also have seen from Earth are those that are exactly the same distance from Earth and your new location; this forms a plane. Anything not within a few tens of millions of light years of this plane will not look the same at all; stars would be born or die, galaxies would move relative to each other, collide or fall into black holes or whatever.

Everything else in the universe would be novel. You would either be looking at things hundreds of millions to billions of years in the past, or hundreds of million to billions of years in the future, from the perspective of Earth. And thus, you would be looking at things that have never been seen from Earth (by humans at least).

Even the largest scale objects in the galaxy would be hopelessly confused over these kinds of time scales. Furthermore, what exactly are the chances that anything large enough to be seen across the universe is aligned with this plane? I can't do that sort of calculation, but the universe is vast, and even a plane 100 million lights years across would only contain a fraction of the objects we have cataloged in the universe. And our catalog is itself a nearly infinitesimal fraction of the objects actually out there in the universe.


I conclude that even with the best star maps produced, and the most powerful supercomputers available, there simply isn't enough similarity between the sky you are looking at and the sky you know from Earth to make any sort of comparisons. That will itself tell you that you are very, very far away, I suppose.

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$ Commented Dec 8, 2017 at 4:53
  • 3
    $\begingroup$ @Fattie Whoops! Someone didn't read the whole answer! $\endgroup$
    – kingledion
    Commented May 4, 2018 at 11:57
  • 3
    $\begingroup$ @Fattie I think you're taking "Travel through space is travel through time" too literally here. When we look 6 billion LY away we see it as it was 6 billion years ago because its light takes 6 billion years to reach us. Stars move. Stars die. New ones are created. We see images from the past in the sky. Somebody 6 billion LY away looking at "Earth" will see a nebula because the Solar System hadn't formed 6 billion years ago. If they jumped to "Earth" instantly they'd miss because the nebula would have moved. It also didn't exist anymore, it had collapsed to form the Sun, Earth and planets. $\endgroup$
    – Schwern
    Commented May 4, 2018 at 18:20
  • 1
    $\begingroup$ @Fattie In short, a map of the cosmos made from Earth is 6 billion years out of date if you instantly jump 6 billion light years away. $\endgroup$
    – Schwern
    Commented May 4, 2018 at 18:21
  • 1
    $\begingroup$ @Schwern, you make an outstanding point. (I perhaps foolishly assumed that in the Universe setting, humans can commonly FT, and hence maps show " 'present' " (probably comoving) positions. If it's a one-off portal you're totally correct.) $\endgroup$
    – Fattie
    Commented May 5, 2018 at 14:39

I waited a whole minute for OP to answer my clarifying question about continuous signalling through the portal. :-) I assume the answer is "yes" now.

Hard yes.

Set up a very-long-baseline interferometer (VLBI) with one aperture on each end of the portal. (Congratulations. You have now made the largest telescope accessible by humanity.) Real VLBIs don't actually need continuous real-time sampling from all participating telescopes. Data can be recorded (at very high speed) and the resulting interferometry done by combining datasets. So my question about continuous signalling through the portal suggests a tougher requirement than is actually needed.

For each sufficiently large $z$ (redshift) as measured from one planet, you will find one arc on the celestial sphere of each planet where the interferometric data has persistently large cross-correlation. This arc corresponds to the directions along which the two planet's past light cones intersect (at the right $z$ as measured from the one planet). That is, along that arc, both telescopes are watching light emitted by the same process the same time-of-flight away from(see below) the two telescopes. If you move off that arc to one side, the time-of-flight of apparent coincident events to both telescopes increases or the time-of-flight to both telescopes decreases, so if there are coincident events, they do not have the selected $z$.(see below)

These arcs can be plotted to make a "bullseye" in the sky. This pattern is centered on the direction towards the latest time (smallest $z$) event that has (or, had) both planets in its future light cone. One could point at the center of the pattern and claim that the other end of the portal is "that way". If the other end is space-like separated from this end, then that claim is hampered by a coherent choice of coordinate system. (It would be more accurate to say that the light from events at a certain $z$ shift in "that" direction travelled in opposite directions to arrive at the telescopes at each end of the transport system. However, during the astoundingly long time it has taken the light to make the journey to each planet, both planets, as well as the light producing object(s) have moved substantially, so where the other end of the portal appears to be (at the bullseye), where the other end of the portal is "now" (whatever that means in the absence of a universal coordinate system), and where you would have to shine a light so that its photons would eventually (maybe) strike the planet of the other portal are wildly different and not practically useful.)

The pattern of the arcs is sufficient to tell you the distance in space and time from each to the latest time (smallest $z$) event that has both planets in its future light cone. As an easy to work out example: If the temporal shift is nearly zero and the spatial shift is nearly zero the patterns are concentric circles. Decreasing $z$ rings shrink down to the direction pointing at the other end of the portal. The planet in the future has slightly larger $z$ shifts than the planet in the past to coincident events.

The $z$ (redshift) of the light from apparently coincident processes along the arcs of fixed $z$ will tell you how far back in time along the cone you have to go to reach the intersection with the other light cone. This is sufficient information to recover the time and spatial shifts.

If the separation in space or time is large, there is a reasonable chance that other galaxies (or other large structures) could appear to lie on the arcs. As a consequence, there could be gravitational lensing making the "an arc" a simplification of the reality of "a narrow-ish band with several complicating micro-features."

Nevertheless, after a few months of observations, one should be able to establish rather sharply where/when is the other end of the portal.

Edit 20171208 13:50 UTC

The text

That is, along that arc, both telescopes are watching light emitted by the same process the same time-of-flight away from(see below) the two telescopes. If you move off that arc to one side, the time-of-flight of apparent coincident events to both telescopes increases or the time-of-flight to both telescopes decreases, so if there are coincident events, they do not have the selected $z$.(see below)

has the cart on the wrong side of the horse. The correct phrasing is

That is, along that arc, both telescopes are watching light emitted by the same process with approximately the same arrival time (say, within a month) to the two telescopes. If you move the event with a space-like separation the two arrival times change oppositely -- the event is observed earlier at one end and later at the other end. If you move the event with a time-like separation, the two arrival times change together, both becoming later or earlier together.

Note that this is an approximate coincidence detection measurement, not an interferometric measurement. The most useful fact about an event is its absolute magnitude, its spectrum, and its $z$. Coincident events have approximately matching absolute magnitude and spectrum.

Further: There are several types of events we could watch for, many of which are susceptible to whole sky surveillance.

It is helpful to know that 6 Gigalight years, 6Gly corresponds to $z ~ 1.5$. (This and all Gly measurements below are in comoving coordinates.)

  • GRBs : BATSE DISCLA data's BD2 sample has about 4500 events with about 1400 quality events ($0.1 < z < 6.5$, or 0.6 to 27 Gly) from a 2 year full sky survey using 1980s technology. See Schmidt, 1999. This gives 50-ish candidate events per month for coincidence detection.
  • Supernovae : IAU Circulars have reported 6264 supernovae this year. This data is collected and summarized here. The range in $z$ for those with measured $z$ (only about 20% of the events) is 0.000133 to 0.915 (0.008 to 10 Gly). Observing supernovae to $z \approx 1.75$ (to 16 Gly) is currently feasible. This gives 500-ish candidate events per month for coincidence detection.
  • Type Ia Supernovae : The Sloan Digital Sky Survey (SDSS) in a 300 square degree area (about 2% of the sky) found 130 SN Ia events in 2005 and 197 in 2006 giving a dozen-ish standard candle (i.e., very well characterized absolute magnitude) candidates per month, or 100-200 such events per month in the whole sky.
  • Quasars : The 2000-2008 SDSS-I and -II surveys observed 100,000 quasars. Subsequent surveys (to the present) have cataloged another 100,000. These have $z$ from 0.056 to 7.085 (0.3 to 28 Gly). This suggests an observation rate of 1000-ish objects per month. Quasars are variables with time scales of hours to months. These would be the first candidates on this list where correlating variations in brightness would be the measurement, rather than just recording coordinate and spectral data for the short event itself.
  • Quasars (again) : The International Celestial Reference System is mostly based on quasars, with measured $z$ up to 4.301 (24 Gly). Many are $1 < z < 3$ (11 to 21 Gly). Consequently, several of these will be in the intersection of the past light cones of two objects with spatial separation 6 Gly and not more than a few Gy time separation.
  • et c. : Turn-on and turn-off events for non-quasar AGNs, and non-EM detections, including neutrino and gravity wave astronomy. LIGO and Virgo have so far reported 4-ish events per year (at distances of 0.13 to 1.5 Gly). Conveniently, the universe is largely transparent to gravity and neutrinos, so interferometry is automatically feasible for these.

So these are the events to measure. What do you do with the measurements? Pick your favourite cosmological spacetime model, for instance FLRW. Call the two portal endpoints "A" and "B". The spectra of events observed at A are compared with spectra for events observed at B. Hough transform matching pairs onto the parameter space of lightcones in your spacetime model. Mismatched events will be scattered over this parameter space. Matched events will lie on/near the surface of intersection of the past lightcones of the ends of the portal.

So far, this has not describe an interferometric technique. However, interferometry for events not "on the line" between the two planets is feasible -- such off-axis events are from more-or-less one side of the event, so coherence increases as the event moves off the axis. Thus, fine-tuning the candidate spectral matches by cross-correlation of short time scale intensity fluctuations, reduces the false matches used to populate the parameter space. (That is, we put less noise in the Hough estimate of the surface of apparent coincidence.)

If the time shift is a bit more -3 Gy or a bit less than 12Gy, then the two past lightcones intersect on a surface that includes the highest density of events listed above, with $z < 1$, from whichever endpoint is earlier. For time shifts outside of this range, the past light cones do not intersect (except at the Big Bang). For time shifts between these, least $z$ for a coincidence decreases to 0.25 for zero time shift. These numbers help us characterize how likely a coincidence event is to be observed during a particular observing window.

As long as the past lightcones intersect, we may observe a coincident event. To simplify calculation, let's pretend events are uniformly distributed on each lightcone. Lightcones extend about 13 Gy into the past. Every month the light cone sweeps through about 1 part in $10^{11}$ of past spacetime of the planet. We distribute 10000 events in that volume, so the probability that none of these events is in the intersection of the lightcones is $1 - (1 - 10^{-11})^{10000} = 99.999990\dots \%$. After a year, $99.99988\dots\%$ chance of no coincidence. This seems hopeless, doesn't it? It's not as bad as that, though, because observed events are not uniformly distributed in $z$. Instead, we are roughly 10000-times more likely to observe an event with $z < 2$ that an event with $z > 6$. (Look at the sources above, $z > 6$ is a once-per-year rarity (60 quasars over 60 years). $z < 2$ is a 10s-100s per day event.) Also, with a separation of only 6 Gly, the $z$ for events "between" the planets are less than 0.75. Consequently, we're scattering 10000-times as many events in half as much light cone. With this adjustment, the odds of no coincidence per month are 90%. The odds of no coincidence in the first year are 28%. So, every few months, we expect to get a new coincident event to update our Hough transform. This is roughly equivalent to our current state of the art in gravitational astronomy - a reportable event every few months.

So my "from the hip" estimate of how long to get space and time shifts was off by an order of magnitude. It will take a few years to get a sharp result. I'm not unhappy with the quality of that estimate.

  • $\begingroup$ Best answer. Why calculate when you can measure? $\endgroup$
    – senderle
    Commented Dec 5, 2017 at 20:21
  • 1
    $\begingroup$ I concede that given the ability to continuously communicate through the portal, this is the correct answer. $\endgroup$
    – kingledion
    Commented Dec 6, 2017 at 23:33
  • 4
    $\begingroup$ Very nice idea, but it has a problem. A galaxy consists of a lot of stars, spread over tens of thousands ly, and Earth and astronauts see it from different angles. There would not be cross correlation from galaxy. You need small bright point-like source - quasar or supernova. But single object like this, watched for a year has about 10-9 chance to be the suitable one. So you have to watch a billion of specific objects continuously for a year. Instead you can look at positions of some millions of them - that gives you a picture of large scale structure of the universe and that's enough. $\endgroup$
    – Vashu
    Commented Dec 7, 2017 at 21:57
  • $\begingroup$ Missed a zero. Actually it is more like 10^-10 chance, and 10 billions objects to watch - so it is even worse. $\endgroup$
    – Vashu
    Commented Dec 7, 2017 at 23:09
  • $\begingroup$ @Vashu: Whole sky surveys have been feasible for decades. Observed rates: GRBs: the BD2 sample of BATSE DISCLA data has about 1400 quality events (0.1 < z < 6.5) from a 2 year full sky survey using 1980s technology. See (Schmidt, 1999. In the first seven days of December, there have been 38 observed supernovae (see rochesterastronomy.org/snimages/sndate.html ). Observing supernovae to z ~= 1.75 (to 16 Gly in comoving coordinates) is currently feasible (arxiv.org/abs/astro-ph/0701584 ). Your pessimistic rates are fiction. $\endgroup$ Commented Dec 8, 2017 at 5:47

This would be hard. You need to look for objects that are:

  1. Detectable from 6 billion ly. That excludes, for example, neutron stars, as we cannot detect old pulsar even outside of our galaxy.

  2. Stable enough to remain recognizable if we would look at it 6 billion years ago. That excludes quasars, that are basically extremely bright accretion disks around giant black holes. Accretion disk does not "hold" any memory - it can easily become much brighter or dimmer, depending on amount of incoming matter, it does not hold matter for long - it gets either consumed by black hole or thrown out.

Also that excludes galactic clusters - in 6 billion years galaxy's can travel some tens millions ly, which would make cluster appearance pretty unrecognizable.

We cannot look at background-radiation anisotropy since it's picture would differ unpredictably for a region so far away.

What we can do is to look for large-scale structure of the cosmos. Great walls and superclusters, Large quasar group, supervoids. With this we can narrow region of search to some hundreds of millions ly, locate Virgo Cluster, our galaxy, and then look for the Sun.

Those large-scale structures were born from fluctuation of density that happened at the very beginning of universe so they are long lived by definition. They are hundreds and billions ly across so peculiar movement of galaxies does not change them much(even 1000 km/s for 6G years is only 20M ly). They would look a little different to our astronauts because of distortions caused by expansion of the universe but this is easy to adjust for.

enter image description here

  • 6
    $\begingroup$ You're still looking for the Virgo cluster of 6 billion years ago - and have no decent way of identifying it. As for these superstructure, remember that we see a young side nearby and an old side far away. From another perspective, this won't be the same. $\endgroup$
    – MSalters
    Commented Dec 4, 2017 at 2:29
  • 8
    $\begingroup$ The large scale structures of the universe would not only be of a different age when we observed them, but different parts of them would be different ages relative to each other. Even if the Virgo Supercluster is equidistant from the Earth and our new point so that we are seeing the same 'time' of it, if we are on the opposite side of it from Earth, then, due to the object's size of 100 million ly, the parts nearer to us are 200 million years younger than what we would have seen form Earth and the far edge is 200 million years older. $\endgroup$
    – kingledion
    Commented Dec 4, 2017 at 4:01
  • 9
    $\begingroup$ @kingledion Large-scale structure of the cosmos are so large it does not matter that we see parts of it at different times. Obviously we must account for expansion of the universe, but that is trivial. If we are talking about some great wall, billions ly across then peculiar motion's of galaxies over billions of years are a couple orders of magnitude less than the size of the structure. Without Hubble expansion(which is easy to account for) it is basically static. $\endgroup$
    – Vashu
    Commented Dec 4, 2017 at 5:21
  • 11
    $\begingroup$ Nice thought, but 6 billion years ago the star which is our Sun didn't exist. Our Sun is about 4.6 billion years old. Travelling to a point 6 billion light years away from "here" would effectively have you looking at light emitted 6 billion years ago, when the Sun didn't exist, so there'd be no way to find it. Also, six billion years ago the Sun wasn't "here" - the dust cloud from which it formed might not have even existed, and was certainly a long way away from "here". $\endgroup$ Commented Dec 4, 2017 at 12:15
  • 2
    $\begingroup$ @Eth a Large Quasar Group is different from a singular quasar. It's current recorded measuements are a widtch that varried from 1.4 billion light years up to 4 billion light years. $\endgroup$
    – PCSgtL
    Commented Dec 5, 2017 at 14:03

Six gigalightyears is halfway to the edge of the universe (13.82 billion light years).

That would create a very large obstacle for astronauts looking to find their way: they would be looking at a night sky about 6 billion years younger in the direction opposite their travel and 6 billion years older in the direction of travel.

We map the sky on Earth the way we see it. Many of the brighter stars have lifetimes of only a few million years, so they would either be long gone or not yet born, or at different phases of their lives (red giants, novae, neutron stars). Even galaxies may look very different over such a large distance. For example, at 4.5 only billion years old, the sun would not even exist in the night sky of the other world, 6 billion light years away.

With back-and-forth travel to Earth available, it might be possible that something in the method of travel provides a rough idea of direction and distance. With that extra clue, you could start drawing inferences about what is what and eventually come up with a guess as to overall location.

However, you could quickly build a new star map and determine where you are in the new night sky (not connected to the bigger picture).

  • 18
    $\begingroup$ I might also add that if we hand-wave relativity and say that they've been transported INSTANTLY that far away, then their cosmological horizon is now different, insofar as there has to be a percentage of the sky that now shows parts of the universe that are currently beyond the Earth's cosmological horizon and therefore outside the observable universe as we know it. That means that not only do the known stars and galaxies look different, but there's a whole bunch out there that we've never seen before as well to consider. $\endgroup$
    – Tim B II
    Commented Dec 3, 2017 at 22:45
  • 10
    $\begingroup$ I thought the edge of the viewably universe is 45 billion, whats 13.82 reffering to. $\endgroup$
    – Necessity
    Commented Dec 4, 2017 at 0:42
  • 24
    $\begingroup$ The 45 billion distance is due to the fact that the universe is expanding as well as getting older. The oldest things are 13b years old, but the farthest are closer to 42b. $\endgroup$ Commented Dec 4, 2017 at 1:05
  • 15
    $\begingroup$ @Ucenna My understanding is that there is no center of the universe. No matter where you are in it, everything around you is expanding away from that point. $\endgroup$ Commented Dec 4, 2017 at 6:10
  • 2
    $\begingroup$ Adding to @WayneConrad's comment above, metric expansion of space appears relevant. $\endgroup$
    – user
    Commented Dec 4, 2017 at 12:17

In addition to James McLellan's good answer, I wanted to point out that while figuring out location is very, very difficult, figuring out the local time* might not be. The CMB temperature should still be a pretty good indicator of "when" the astronomers arrived (and there are other global quantities that shouldn't depend on position which evolve very regularly in time that they could measure as well). If the astronomers traveled backwards or forwards 6 billion years, the CMB would be considerably hotter or colder than the 2.7 kelvin we measure at Earth - I'm not familiar enough to know how precise they could get with this, but they could certain determine any big shifts in local time*.

*Local time meaning the "proper time", that is, the time elapsed since the big bang in that region of space. As you may know, defining simultaneous times in cosmology is pretty tricky.


The simple answer to the question is: no.

This problem is wonderfully depicted in the otherwise bearly tolerable movie Lost in Space. Sabotage to the the hyperdrive and traditional Hollywood circumstances send the Jupiter II sailing through space to unknown locations where their "star charts" are useless.

Why are they useless? Because it's the biggest 3D jigsaw puzzle of all time. Enough information about enough stellar phenomena must be stored in a database with enough computational power to crank through what is fundamentally an infinite combination of positions within the galaxy.

It is helped somewhat with some POV references such as the galactic core and known pulsars, quasars, etc, that are unique enough objects that can yet show up at those distances... but still...

You need time to capture information. A quick snapshot of the sky to let the computer start chewing while you capture even more detailed snapshots, and even more detailed snapshots... with the downside that with every enhancement of resolution you are exponentially increasing the amount of computational time needed to find your position in the galaxy.

The more complicated answer is: yes.

Given a good enough database and time, you can find your location. Except...

But the even more complicated answer is: maybe.

My knee-jerk reaction to "can you peer to the diametrically opposite side of the galaxy?" is "no." I suspect no technology can peer through the galactic core to the other side. (super massive black hole eating all the electromagnetic radiation...). Therefore, without expansion throughout enough of the galaxy to give you the vantage points for deep peering, there are spots in the galaxy you simply no nothing about. Use a portal to one of those locations and, basically, it's impossible to know where you are.

But, if you had the tech and the expansion time to peer around and see all the corners for mapping purposes, then the six billion lightyear jump is no longer impressive. You have likely expanded more than that just to see the galaxy's rear end... because obviously humans are at the front end of the galaxy, dontchaknow.

  • 3
    $\begingroup$ Umm... six billion light years is about 60,000 times the size of our galaxy. Why would one need to "expand more than that" just to map that galaxy? $\endgroup$ Commented Dec 4, 2017 at 0:24
  • $\begingroup$ Because it's the biggest 3D jigsaw puzzle of all time Quite. The 90s adventure game Starship Titanic featured a puzzle at the end that was basically that. $\endgroup$ Commented Dec 4, 2017 at 1:07
  • 2
    $\begingroup$ That black hole lenses - you can see further in that direction. $\endgroup$
    – MSalters
    Commented Dec 4, 2017 at 2:30
  • $\begingroup$ In Lost In Space, they did manage to download more detailed star charts, but even then, that wasn't that helpful. $\endgroup$
    – ackmondual
    Commented Dec 4, 2017 at 3:00
  • 1
    $\begingroup$ @HenningMakholm, I wrote my answer without caring about the distance. It was an oversight on my part. However, greater distance simply exacerbates the problems I indicate. $\endgroup$
    – JBH
    Commented Dec 4, 2017 at 5:28

I think the astronauts would take a two step process. The first step is easy. They need to get a background-radiation satellite mapping the sky on the other side. You might try grabbing an existing deactivated satellite like WMAP or make another. Either way, get one on the other side and have it start mapping.

While it's mapping, run a few experiments to make sure basic physics assumptions don't fall through. Toss a few atomic clocks through and try to find out whether they tick at the same rate as they do on the near side of the portal. Remember, this is new science, so you can never be too careful.

When you're done, you should have a map like this:


This is the output of WMAP after 9 years. As many local effects have been filtered out as possible (such as the effect of our own sun's movement through space), leaving only the anisotropy that we believe is associated with the cosmic background radiation.

This should be matchable to the readings taken on the other side of the portal. There might need to be some adjustments if timetravel was involved.

Once you have this, you have a solid orientation anchor. You know what is up, down, left, and right. Next, I'd use a surveying approach. Take a look for quasars. We should be able to find more than enough of them. Once you have a good set of readings, you can start using surveying techniques to find the best match for the angles that we see on both sides of the portal.

Once you have a solid match, along with those angles, now we should be able to figure out where we are, with respect to the quasars that can be seen from both places. Quasars are separated by several billion light years, so the angles should meaningful. If you were to travel a few trillion light years, it might be harder.

  • 3
    $\begingroup$ Quasars are basically extremely bright accretion disks around giant black holes. Accretion disk does not "hold" any memory - it can easily become much brighter or dimmer, depending on amount of incoming matter, it does not hold matter for long - it gets either consumed by black hole or thrown out. So some quasar 6 billion years ago would look unpredictably different. $\endgroup$
    – Vashu
    Commented Dec 4, 2017 at 0:07
  • 9
    $\begingroup$ Background-radiation anisotropy comes from adiabatic density perturbations in some region of space and late time anisotropy scattering. So this picture in a place 6 billion years away would look completely different, except may be for a narrow ring(ring that comes from intersection of 2 spheres). Maybe you could correlate data for that ring, but I am not sure about it. $\endgroup$
    – Vashu
    Commented Dec 4, 2017 at 0:11
  • $\begingroup$ Though I guess we know quasar positions pretty well(though they would move a lot in 6 by) so there is possibility that we identify them by relative positions. $\endgroup$
    – Vashu
    Commented Dec 4, 2017 at 3:34
  • $\begingroup$ The microwave radiation that we observe today originated from a locus of points that form a hollow sphere centered on where the Earth is today. That light started out at the moment the universe became transparent, a few hundred thousand years after the Big Bang, or about 13.8 billion years ago, so the sphere has a radius of that many light years. There's absolutely no reason to think that the fine details of the radiation converging on a point 6 billion light years away, which originated from an almost completely disjoint sphere, would look anything like what we see here. $\endgroup$
    – Dave Tweed
    Commented Dec 7, 2017 at 2:28
  • $\begingroup$ @Vashu That ring is exactly what they'd be looking for. We have already looked for rings in the CMB in order to determine if the universe is a closed surface below a certain size, and have convinced ourselves it is not. The same technique can be used to determine roughly when and where another location is; the shape and location of the identical/similar rings gives us a relative location, so long as the CMB overlaps. Possibly it would give a family of possible time/space coords, which you could then use to double check using other methods (great voids, etc) $\endgroup$
    – Yakk
    Commented Dec 8, 2017 at 14:09

Hard Yes

Some facts and observations from other answers:

  • Observable universe is 93 Gly
  • There would be a plane where equidistant stars are seen at the same 'time' 3 Gly away from both locations
  • Cosmic background radiation can be reliably used to establish orientation
  • There are few very large structures that don't change too much over the time scale we're interested in (between 0-6 billion years)

Given enough time, a survey will find both the large structures and the plane where the stars are the same from both sides. Which is plenty of information to determine the exact location of Earth


It's trivial:

our observable universe, in fact contains

only a few very large structures,

which are easily identifiable.

You'd just look for and identify the sloan great wall, the Bootës void, and the 3 or 4 biggest superclusters.

(Obviously, this would take incredible telescopes and a few years.)

{Note that we already have "incredible telescopes". Read up on Europe's GAIA space scope, which is pure "sci-fi".}

Vashu has already included in his answer and overall map of our observable universe, which shows how obvious it is.

Note. OP's question was not clear if the travelers were "instantly" transported in an FTL sense. If the travelers are instantly transported, then as I explain it's trivial.

If OP means moving billions of years in time, the question is meaningless. (There's no way whatsoever to predict where everything was/will be over huge time periods.)

  • 7
    $\begingroup$ In a relativistic universe, moving in space is equivalent to moving in time. There is no global "now". $\endgroup$
    – Mark
    Commented Dec 4, 2017 at 23:04
  • $\begingroup$ @Mark - newsflash - there's actually no FTL :) Of course, obviously, most of the QA on this site are totally unrealistic, including the many dealing with FTL issues. $\endgroup$
    – Fattie
    Commented May 4, 2018 at 11:10

As long as the astronaut has a map of the sky from Earth where stellar bodies are grouped by distance (like on onion peels), then geometry suggests that there must be at least one point, or more likely a ring in the alien sky that looks rather similar, if not identical to a ring of equal size from the Earth sky. Get the distance of the celestial bodies in the ring and you have the half distance from Earth (3bn light years in this example).

AN algorithm to find rings with certain characteristics in the sky has been used to study and identify echoes in the background radiation if I remember correctly.

  • $\begingroup$ This should also allow to establish the direction as the orthogonal direction to the plane that contains the ring, passing through Earth and on the other side of the plane relative to Earth. $\endgroup$
    – NofP
    Commented Dec 4, 2017 at 9:29
  • $\begingroup$ Probably even simpler, considering that every celestial body on the plane that is half-way between Earth and the new location is half-way between the two, hence it should look the same. Further, there should be no "similar-looking" celestial body which is closer than the half-distance. one could get an approximate upper bound on the minimum distance by trying to identifying objects in the alien sky, clustered by increasing distance from Earth. $\endgroup$
    – NofP
    Commented Dec 4, 2017 at 23:20
  • $\begingroup$ The hard part is having the computational power to churn through all the possible rings among the observable objects in the sky... Fortunately the process would be parallelizeable, so some of the up-and-coming new computational technologies might make it feasible. $\endgroup$
    – Perkins
    Commented Dec 7, 2017 at 0:06
  • $\begingroup$ The OP specified that the wormhole could make them travel through time too. So they might be many years in the future or in the past. Which means that the ring that looks similar will not be halfway. It would be closer to the ones that are seeing it from the "past", if I am visualizing it the right way ^^ $\endgroup$ Commented May 22, 2018 at 13:47

Well, assuming hefty doses of handwavium so the astronauts survive the radiation inherent in a wormhole, I believe it would be possible, although it might take some time.

The secret is a total sky galaxy survey, backed up with a corresponding survey from earth such as has been/is being done by HST, Spitzer, and Chandra. Once you positively identify four galaxy matches with their red shifts, you can pinpoint the milky way (even if it's hidden behind another galaxy) and determine how much time has elapsed since you left home to transit the wormhole.

I'll let an astronomer do the math for this one.

  • 1
    $\begingroup$ Yes, once you positively identify four galaxy matches, you know your location. The problem is identifying those four matches: a 6 giga-lightyear shift in position means an up-to-six-billion-year shift in what time you're seeing things from. Do you know what any of those galaxies looked like six billion years ago? Or six billion years from now? $\endgroup$
    – Mark
    Commented Dec 4, 2017 at 22:55

Yeah they would. They would just have to look for literally anything familiar in the sky. Once you find something you recognize you can then estimate your distance from earth using triangles. For example, you're on planet X. You see a familiar planet Y that is around 15 million light years away from your planet at 30 degrees from your planet - just assume everything is on the same plane, all that does is simplify the maths idea. You then walk through the portal. As luck would have it you can see planet Y and you recognize it, which means that the light is probably the same age, since you literally just recognized it. Now all you have to do is figure out what angle it is away from you, graph all your points and then find the missing distance from your new location to planet X.

There are also light dating techniques you can get into as well so if can't recognize anything you could start drawing the universe from your position and eventually you might find an area that looks like what you'd expect an area you've observed from earth would look like at the time reflected by the difference in the age of the light.

Admittedly, both of these techniques are somewhat iffy. It might be easier to just figure out how the warm hole works.

Or... or... don't read this if you're easily triggered... or, you can just try to detect gravity waves and hope you detect a wave you've already detected on earth and then try to use that to estimate where you are. You'd have to assume that waves that look the same are the same, but it could be done.


You may be able to find out where the astronauts are from the wormhole itself.

Suppose in your world that the wormhole has a direction and a relative time that a travel experiences while passing though. Suppose it take a X length of time to pass. If your astronauts could travel back and forth and accurately measure this time gap while in the worm hole they could relate this to the actual distance. Suppose 1 minute of wormhole travel is equal to a billion light years of actual distance or something like that. You could even measure the time in nano seconds if you want it to feel more instantaneous. If your wormhole is 2D you could just align the orientation of the entrance to the wormhole to get the direction your astronauts would travel. With the distance traveled and a direction your astronauts could get a good idea where in the universe the other side is.

If you have a spherical 3D wormhole it might be more difficult to get the direction. Perhaps you cold play with symmetry that the wormhole opens up at a symmetric point on the opposite side of the universe like if I dug a hole though the earth in a straight line I would end up at a specific point on the other side.

  • 3
    $\begingroup$ Any wormhole based on actual physics has a transit time of 0. $\endgroup$
    – Mark
    Commented Dec 4, 2017 at 22:56
  • $\begingroup$ And if the orientation is dependend on the other end you might want to stop the movment of earth, sun, milkyway and such first. Or jump through really fast. ;-) $\endgroup$
    – Henning M.
    Commented Dec 5, 2017 at 16:26

The astronaut couldn't find out where he is or when, if he doesn't have data from the night sky on earth. He would need more than a Polaroid of the sky at night^^. I guess NASA won't expect that so they wouldn't put such data into space ships.

But if they have a data link to earth (Think Stargate, where radio waves could go both ways, matter only one way.) and a strong enough telescope at their end of the portal, they could: Find a stars, a galaxy or a black holes that are the same distance from them and from earth. The light going out from them would reach the earth and the portal exit at the same time. If they find one, it would be easy to get your position relative to Earth by looking at how it is rotated. For example if you see the galaxy from earth at a 45° rotation. And from the portal at 135° you had a triangle with a right angle (a²+b²=c² and a=b). If they find more than one it would be even better. If they don't find anything at exactly the same distance with earth. Maybe something recognizable even if it is 2 billion years wrong. And than you could find the Milky Way and look for our sun. (There shouldn't be anything to see, since light of our sun has only travelled 4,5 billion light years, cause our sun is 4,5billion years old. But we should be able to see his sun, if it is older than 6 billion years.) Or some other light source that was old enough.

If the wormhole also travels through time*, it would still be possible to find something that has send the light at the same time out, and reached the astronaut and earth at roughly the same time. And you could find out that happened, because the calculated distance of the wormhole exit and earth wouldn't match. (You would need to look at something near us that is old enough that light reached the astronaut.)

*Well as long as the wormhole exit has light from stars come to it that also travelled to earth. check out https://www.space.com/33005-where-is-the-universes-edge-op-ed.html So it would be impossible to find out for the astronaut, if the wormhole would send him over 26 billion light years away (without time travel).

So I would say it is possible, but if he is alone with no communication, he would need years, strong computers and a good (radio) telescope. NASA would probably need also years just to get the necessary equipment through the wormhole.


Yes, but not quickly and requiring a few assumptions. Essentially I'm expanding on other people's answers with more optimistic assumptions.

The important assumption is that there exist enough uniquely identifiable objects in the sky, that is, objects that can be identified without Earth-centric context. I'm thinking of objects like distinctive pulsars, binary systems with distinctive variation, easily identified exoplanet systems, etc. With enough of these objects we would have a map of known objects in known locations at a known time.

As pointed out in other answer, a system so far away only has a small plane of objects in similar enough states to be identical but that's fine. At the other end of the portal we survey the sky and build the same kind of database. Providing there's sufficient points in the first database we'd expect to find at least 1 point in common, which sets a common point of reference. Given 1 reference both planets are on a the surface of a sphere that radius from that reference, a good second common reference reduces it to the circumference of a circle and a third pin points.

The range of objects can be expanding if we can apply some stellar mechanics... if we can estimate the size of a supernova from the star that causes it, or the size of the star from a stellar remnant, we can widen the range of possible matching candidates at the expense of certainty on each one.

I'd like to think that with something comparable to Hubble we could try to match some galaxies. We don't need to match galaxies that are considerably off the equidistant plane, even a small distance off that shouldn't be too different.


The first step would be look at the local stars, maybe take a few spectrographs and decide you are not local, or able to recognise any near by stars.

Then look for pulsars / quasars and decide you still cannot map any of them, cause over 6 GYs they either are not active or out of sight.

The final option which may eventually tell you where you are, over a long period, is to very carefully study the standard candle supernova events. Eventually some correlation may be observed over events which happen half way between Earth and this new world. I am going to guess you would likely need to compare records over a few 100 years to get a really good set of events that you can confidently say are the same on both sides.

  • 1
    $\begingroup$ The problem lies in the correlation. As far as I know, the Type Ia supernovae are essentially indistinguishable -- that's what makes them so useful for measuring distance. $\endgroup$
    – Mark
    Commented Dec 4, 2017 at 23:03

You must log in to answer this question.

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