The practical limitation isn't just magnitude, but resolving power. Atmospheric disturbance makes it difficult to resolve anything smaller than a meter from orbit, and the diffraction limit means the further are away from the thing you want to observe, the wider your telescope needs to be. Since the distortion effects near a black hole are near-asymptotic, the features we want to see have very small subtended angles, and we need much higher resolving power to "see" recognizable features.
Stars? Definitely feasible. Planets? Much harder. People? Certainly not with available optics.
To be technical, you could still see an object that is smaller to than your resolving power (this happens essentially every time we observe individual stars). The issue is what is called the "confusion limit" - the point at which you can't be confident that what appears to be a source is not just the summed emission from a bunch of fainter background objects.
But you are right that increased resolution is key. The higher resolution helps you beat the confusion limit by working to resolve the faint background sources into individual objects, so you can separate them.
So, if you wanted to be certain that what you were seeing was the refracted light from Earth, you would need the high resolution to be sure that what you were seeing was actually that reflection, and not just a superposition of fainter background sources.
yes but if we manage to do everything (including unwarping the earth's reflection in the image) we would only see how our planet looked like 6 light years ago (3 for the light to arrive to the black hole and 3 to go back) and we won't be able to discern anything but the earth's surface, is that right?
oh -- yeah seeing what people are doing on Earth via a black hole is totally out of the question technologically. that's pretty much the conclusion of the article.
You don't want the light reflected by 180 degrees - it needs to be reflected where Earth is going to be; a much harder problem. For the easy case quoted in the article of 6 light years, any collimated light would need to be properly directed at the sub-nanoradian (milliarcsecond) level (which is possible -- see sig and [0]).
Interesting to think about, even if it is an incredible observational challenge.
Fun application for the Arecibo radar, if it's still running. Locking into a modulated signal might yield the necessary SNR?
"it needs to be reflected where Earth is going to be; a much harder problem"
The relative velocity of the ancient Earth, black hole, and present earth would also need to coincide in such a way that you'd get a continuous (or, less desirably, periodic) stream of photons making the complete round-trip. Such a coincidence seems pretty unlikely.
Actually it seems pretty inevitable. Light is sprayed across the black hole from Earth; some of it ends up in every possible trajectory; some ends up in the critical 'orbit window' to be reflected back to where earth is going to be.
Actually it makes the problem easier because the angle is less than 180 degrees which means the lensing happens further away from the blackhole with less noise. By my quick googling and back o' the napkin calculations using MACHO-96-BLG-5, we, the Sun and the Milky Way, have traveled about 24 light years in 6,000 years giving an angle of 179.5 degrees! OK, that didn't change much.
I don't know if it's harder to get the light to where earth is going to be. There's still exactly one angle that will work, and we would have to trust to luck either way to send us the right photons. It's not like anyone is aiming.
I had an idea to write a story about this when I was a kid. In my version, planets made of pure mercury were sought by detectives for their potential crime-solving reflections of Earth at various points in history.
This is really off topic, but once - I think I was 13 or 14 - I found a book just lying around somewhere at what passed for my summer camp. I read the whole thing; I guess I must have tried to return it, but I don't remember that part. The book was interesting enough that I wanted to read it again at some point, but I had long since forgotten its name.
This is that book.
I guess I could have found the name of the book by actively asking around, but still - I love the coincidence. Thank you.
This book has come up several times in several recent (unrelated) hacker news threads. I find the prescience of the privacy concerns (and how people react to never ever having privacy again) relevant to today's world of drones and oversharing via social media. And of course it's just got that good Clark science tone.
Does anyone here know why the Einstein cross has 4 of the repeated image instead of being a continuous repetition around the lens object? It seems to me like you should see the magnified image as a sort of circle instead of having a finite number of repetitions.
Is this related to the lines forming a cross you see in pictures of stars?
The lines forming a cross in pictures of stars is a separate phenomenon. That is due to diffraction from the support structures of the secondary mirror.
I'm not so much into physics but based on what I understand - ok, we might get just enough photons reflected back by MACHOs if we build a massive new telescope.
But then how do we separate them from other photons - ie the ones not reflected back? Do photons reflected by MACHOs have a specific signature on their EM wave ??
That is exactly the problem. The author talks about the photons that get turned 180 but the reality of the situation would be similar to seeing our sun reflected in a piece of glass a few thousand light years out (no need for a MACHO). The reflection is there but so is all the background noise, so just like trying to look the reflection on a clear piece of glass, the light mixing would be killer.
the photons "reflected" back would appear at specific locations around the edge of the MACHO, so you'd see expect to see something with Sun-like spectra at that radial location. lots of image processing would be required to un-warp the collected image.
I find it sort of ironic that everyone here thinks this would be awesome, but are completely opposed to hovering predator drones observing our every action 24/7. Obviously it's all a matter of who has access to the data, but essentially the data would be the same.
Even disregarding the resolution problems as brought up by ebilsten, and assuming we have some magical tech that would allow us spy-satellite-like capabilities with this... we'd still be looking at Earth decades in the past.
I for one could not care less if historians centuries in the future want to observe me through a black hole. Its effects on freedom and civil rights is basically nil, unlike predator drones.
In Orson Scott Card's "Pastwatch: The Redemption of Christopher Columbus" (awesome book, btw! I don't think he's planning on writing the remaining 2 in the trilogy, though :() they have machines that do exactly this, letting them monitor ancient history up to a decade ago: anyone, anywhere on the planet. The catch is that the builders of the machine actually have access to real-time monitoring, but pretend this isn't the case to avoid public outrage.
The data is not the same - they vary greatly on spatial resolution and latency. Better than 10cm resolution with effectively 0s latency (in the drone case), and ~5,000-12,000km resolution (assuming you can get back only a couple of pixels from a MACHO) with yrs of latency are two very very different things.
Was reading the article, excited by the prospect when I realised, we have a hard enough time identifying planets dozens of light years away, from direct light, that surely the reflected light from a massive object 3000 light years away would be orders of magnitudes more difficult. Perhaps impossible. Theoretically possible yes, but what sort of technology needs to be invented to actually do it?
edit: turns out I should have just finished reading the article.. bigger telescopes!
> If we ever invented FTL travel, even for just probes, this could be done too.
If we had FTL travel, we could outrace light waves, look back, and see events in the past. The we could race back and report what we saw. That's why it's not possible -- it violates some basic physical ideas about causality.
Remember that the speed of light isn't just a speed limit for light waves -- it's a barrier that sorts out temporal causes and effects.
Imagine a flat plane in the time dimension called the present -- on each side of the plane is a cone. The cone in the past encloses possible causes for present events. The cone in the future encloses possible effects from events in the present. Outside the past cone are events that may not produce results in the present. Outside the future cone are effects that cannot have been caused by events in the present. The slope of the cone surfaces is the speed of light.
>>If we had FTL travel, we could outrace light waves, look back, and see events in the past. The we could race back and report what we saw. That's why it's not possible -- it violates some basic physical ideas about causality.
You are confused. You are not actually traveling to the past therefore causality is not a problem here.
I'm confused? Which of us is discussing FTL without understanding its physical meaning?
> You are not actually traveling to the past ...
Locate where I made this claim.
> ... therefore causality is not a problem here.
Causality is certainly a problem. FTL is impossible in the same way that dividing a number by zero is impossible or taking the square root of a negative number is impossible.
Here's the simplest form of the relativistic equation that tells us the amount of time dilation we can expect for a given velocity v:
t' = t * sqrt(1-v^2/c^2)
t = time rate at v = 0
t' = time rate at v
c = speed of light
v = frame velocity
Compute the time dilation for v > c, then get back to me.
FTL means leaving the causality light cone, conventional notions of causality, and therefore all of physics:
Quote: "Because signals and other causal influences cannot travel faster than light (see special relativity and spooky action at a distance), the light cone plays an essential role in defining the concept of causality - for a given event E, the set of events that lie on or inside the past light cone of E would also be the set of all events that could send a signal that would have time to reach E and influence it in some way." [emphasis added]
> Causality is certainly a problem. FTL is impossible in the same way that dividing a number by zero is impossible or taking the square root of a negative number is impossible.
This assertion is false. GR, even without FTL, allows for closed time-like loops, which violate causality. Violations of causality are not logical contradictions like dividing by zero. (Though, of course, there are alternative mathematical systems that allow for dividing by zero. And taking the square root of a negative number is surely not impossible. You just get an Imaginary result.) It's just that a universe in which causality is violated all the time would be a very strange one, and consequently, it is reasonable to assume that if there are, for instance, closed time-like loops in the universe, they are not traversed very often.
Quote: "In the theory of general relativity, the concept of causality is generalized in the most straightforward way: the effect must belong to the future light cone of its cause, even if the spacetime is curved."
Quote: "Weird space-time geometry distortion within general relativity can result in causality violation, but there is no evidence that they are physically possible."
> Quote: "Weird space-time geometry distortion within general relativity can result in causality violation, but there is no evidence that they are physically possible."
And there's no evidence that they are not possible. With no evidence either way, one must remain agnostic.
On the other hand, we do have evidence that closed time-like loops are possible: GR has never been observed to be wrong, and GR predicts their existence under the right circumstances.
It may certainly be the case, that even though closed time-like loops are possible, they do not actually exist in our universe and never will. Paradoxical universes are ruled out by logic alone, and so the space of possible universes without closed time-like loops (or least without humans traversing them) may outnumber the space of possible universes with time-like loops in them, allowing us to make a probabilistic argument against their occurrence in our universe. My transfinite number theory is rusty, however, so YMMV.
> And there's no evidence that they are not possible.
In science, the default precept toward ideas without evidence is the null hypothesis -- they are presumed to be false. This is why Bigfoot is not accepted as a scientific theory -- until it is observed, it doesn't exist.
Quote: "In statistics, the only way of supporting your hypothesis is to refute the null hypothesis. Rather than trying to prove your idea (the alternate hypothesis) right you must show that the null hypothesis is likely to be wrong – you have to ‘refute’ or ‘nullify’ the null hypothesis."
Saying "there's no counterevidence, therefore ..." is not a scientific statement, because it can be said about anything not conclusively falsified, including meadowland fairies and honest politicians.
> With no evidence either way, one must remain agnostic.
Not in science. It's nice to have an open mind, but not to the degree that one's brains spill onto the ground.
This is just untrue. When presented with a hugely precise and successful theory that has never been demonstrated wrong in even the slightest degree, that is certainly evidence (not proof, but evidence) towards whatever claims that theory makes. If you don't understand this, then you don't know what the word "evidence" means.
> This is just untrue. When presented with a hugely precise and successful theory that has never been demonstrated wrong in even the slightest degree, that is certainly evidence (not proof, but evidence) towards whatever claims that theory makes.
Nonsense -- perfect nonsense that the history of science repeatedly falsifies.
1. Science is empirical -- it must be. If it's not empirical, it's not science.
2. GR succeeded with every challenge that was put before it until it was compared to quantum theory, at which point it failed. Your position is that, because it had passed any number of difficult tests until then, it should have been assumed to reflect reality in all respects.
But this is not how a scientist approaches theories and experiments. Any aspect of a theory that has not been tested, is assumed to be false until there is evidence to support it.
The ether theory of the 19th century was successful until it failed one crucial test. The Phlogiston theory was successful until it failed one crucial test. The Ptolemaic theory of orbital mechanics seemed correct until Galileo looked through his telescope.
All scientific theories are assumed to be false until there is evidence to support them, then, after evidence supports them, they are assumed to be perpetually falsifiable by new evidence. An unfalsifiable theory is not a scientific theory.
> If you don't understand this, then you don't know what the word "evidence" means.
Science is not law. Learn about science.
> When presented with a hugely precise and successful theory that has never been demonstrated wrong in even the slightest degree ...
You need to learn about quantum theory, which falsifies GR. Start here:
The present enthusiasm for superstring theory is that it might -- might! -- replace both GR and quantum theory, with a more comprehensive theory that would resolve the conflict between the two earlier theories.
> Nonsense -- perfect nonsense that the history of science repeatedly falsifies.
You are completely incorrect in this and everything else you said above. Well, it is true that science is empirical, but it seem that you have profound misconceptions about what "empirical" means. One of the things that it means is that every additional confirming piece of observed evidence for a theory gives extra weight to the probability that the theory is correct, especially when the theory makes correct predictions against previously unobserved phenomena, which GR has done many times.
You also apparently don't understand that theories are not typically purported to be facts per se, but rather they are models. Models that have greater or lesser accuracy. Models are not expected to be perfectly accurate, which is why they are models and not facts. Though in fundamental physics, we do expect a level of accuracy far above what we would expect in almost any other science.
The ether model was proven to be so inaccurate as to no longer be useful. The limitations of Newtonian mechanics were shown by Relativity, but nonetheless, the model is still a hugely important and useful model, and using it correctly is the epitome of good science, despite the fact that we know that in some sense it is "false".
Re your examples, all the examples you give of theories that were eventually proven incorrect are arguments against a strawman, because no one here or anywhere has claimed that any amount of empirical evidence will grant a theory a 100% probability of correctness.
No one has asserted that GR is unfalsifiable, or that it has a 100% probability of being correct. None of this means that we can't have an extremely high confidence level in GR, considering all the empirical gauntlets that it has successfully passed. And when we finally have a successful TOE, GR is almost certainly not going to be scrapped. It will still be the extremely useful and productive model that it has always been. We will just have to be aware of the model's caveats, just as we must with Newtonian Mechanics, which is also one of the most successful scientific models ever constructed.
Additionally, all this is neither here nor there for the original point, which was about your claim that FTL is logically impossible. That is just not the case. FTL is completely within the realm of possibility. It just has serious and problematic consequences that should make us look askance when people claim to have figured out how to make FTL drives or communication channels. They are almost certainly mistaken, but not for the reasons that you claim.
You are forgetting that we could "travel faster than light" using an Alcubierre Drive [1][2], or something similar. i.e. bending space so that our path is now shorter even though in our local space we are not traveling faster than light.
"FTL means leaving the causality light cone, conventional notions of causality, and therefore all of physics:"
While I agree with your conclusion that ftl is equivalent of time travel and breaks causality I think its incorrect to say it breaks all of physics. There are several solutions to general relativity that are valid. http://en.wikipedia.org/wiki/Chronology_protection_conjectur...
The light speed limit comes up elsewhere. As best we know, you cannot transmit information faster than c. Not even with "spooky quantum stuff". There are no shortcuts. Information is just as real as matter or energy--if not more so, see John Wheeler's "it from bit"--and isn't tied to meaning, see Shannon.
Except of course if you have invented FTL travel, you can use it to actually go to the earth in the past. FTL drives are also time machines. There's a saying in Physics. Choose two: FTL, causality, or relativity. You cannot have all three.
Another possibility (perhaps more feasible?) would be for an intelligent alien civilization to observe light / radio transmissions from Earth, record them, and kindly play them back for us.
I've never seen that image of our galaxy's central region, it's fascinating! Is "Sgr A" the black hole, Sagittarius A*? If yes, why would it be bright? I thought it's inactive.
Sgr A is the black hole Sagittarius A* as far as I know. The black hole itself isn't bright. As far as I understand, the swirling gas around it (outside the event horizon) is what's sending out all the light.
By inactive I meant that it's not swallowing anything currently, but I guess if it has attracted a gas cloud this is simply not true. Must be quite a sight. Imagine there's a civilization on one of the central stars (if that's even possible considering the bursts of radiation they must get there) - I wonder what the night sky would look like.
The activity level (in terms of how much material is falling into it) is fairly low, such that it doesn't qualify as being an "active galactic nucleus". However, a gas cloud (called "G2") will be passing close enough to the supermassive black hole that it will be tidally disrupted and part of the cloud will fall into the black hole. This will give us a chance to see enhanced accretion, though still not of the level for it to be considered "active".
Greg Egan had a short story ('The Hundred Light Year Diary') in his collection 'Axiomatic' which dealt with people be able to send messages to the past (and thus becoming aware of their future).
Stars? Definitely feasible. Planets? Much harder. People? Certainly not with available optics.