Astronomy Without A Telescope – Apparent Superluminal Motion


The recent list of Universe Today’s Top 10 Stories of 2010 included the story Faster than Light Pulsars Discovered – which on further reading made it clear that the phenomenon being studied wasn’t exactly moving faster than light.

Anyhow, this prompted me to look up different ways in which apparent superluminal motion might be generated, partly to reassure myself that the bottom hadn’t fallen out of relativity physics and partly to see if these things could be adequately explained in plain English. Here goes…

1) Cause and effect illusions
The faster than light pulsar story is essentially about hypothetical light booms – which are a bit like a sonic booms, where it’s not the sonic boom, but the sound source, that exceeds the speed of sound – so that individual sound pulses merge to form a single shock wave moving at the speed of sound.

Now, whether anything like this really happens with light from pulsars remains a point of debate, but one of the model’s proponents has demonstrated the effect in a laboratory – see this Scientific American blog post.

What you do is to arrange a line of light bulbs which are independently triggered. It’s easy enough to make them fire off in sequence – first 1, then 2, then 3 etc – and you can keep reducing the time delay between each one firing until you have a situation where bulb 2 fires off after bulb 1 in less time than light would need to travel the distance between bulbs 1 and 2. It’s just a trick really – there is no causal connection between the bulbs firing – but it looks as though a sequence of actions (first 1, then 2, then 3 etc) moved faster than light across the row of bulbs. This illusion is an example of apparent superluminal motion.

There are a range of possible scenarios as to why a superluminal Mexican wave of synchrotron radiation might emanate from different point sources around a rapidly rotating neutron star within an intense magnetic field. As long as the emanations from these point sources are not causally connected, this outcome does not violate relativity physics.

2) Making light faster than light
You can produce an apparent superluminal motion of light itself by manipulating its wavelength. If we consider a photon as a wave packet, that wave packet can be stretched linearly so that the leading edge of the wave arrives at its destination faster, since it is pushed ahead of the remainder of the wave – meaning that it travels faster than light.

However, the physical nature of ‘the leading edge of a wave packet’ is not clear. The whole wave packet is equivalent to one photon – and the leading edge of the stretched out wave packet cannot carry any significant information. Indeed, by being stretched out and attenuated, it may become indistinguishable from background noise.

Also this trick requires the light to be moving through a refractive medium, not a vacuum. If you are keen on the technical details, you can make phase velocity or group velocity faster than c (the speed of light in a vacuum) – but not signal velocity. In any case, since information (or the photon as a complete unit) is not moving faster than light, relativity physics is not violated.

3) Getting a kick out of gain media
You can mimic more dramatic superluminal motion through a gain medium where the leading edge of a light pulse stimulates the emission of a new pulse at the far end of the gain medium – as though a light pulse hits one end of a Newton’s Cradle and new pulse is projected out from the other end. If you want to see a laboratory set-up, try here. Although light appears to jump the gap superluminally, in fact it’s a new light pulse emerging at the other end – and still just moving at standard light speed.

Light faster than light. Left: Stretching the waveform of light can make the leading edge of the wave seem to move faster than light. Right: Gain media can act like a Newton's Cradle, making light seem to jump the gap superluminally.

4) The relativistic jet illusion
If an active galaxy, like M87, is pushing out a jet of superheated plasma moving at close to the speed of light – and the jet is roughly aligned with your line of sight from Earth – you can be fooled into thinking its contents are moving faster than light.

If that jet is 5,000 light years long, it should take at least 5,000 years for anything in it to cross that distance of 5,000 light years. A photon emitted by a particle of jet material at point A near the start of the jet really will take 5,000 years to reach you. But meanwhile, the particle of jet material continues moving towards you nearly as fast as that photon. So when the particle emits another photon at point B, a point near the tip of the jet – that second photon will reach your eye in much less than 5,000 years after the first photon, from point A. This will give you the impression that the particle crossed 5,000 light years from points A to B in much less than 5,000 years. But it is just an optical illusion – relativity physics remains unsullied.

5) Unknowable superluminal motion
It is entirely possible that objects beyond the horizon of the observable universe are moving away from our position faster than the speed of light – as a consequence of the universe’s cumulative expansion, which makes distant galaxies appear to move away faster than close galaxies. But since light from hypothetical objects beyond the observable horizon will never reach Earth, their existence is unknowable by direct observation from Earth – and does not represent a violation of relativity physics.

And lastly, not so much unknowable as theoretical is the notion of early cosmic inflation, which also involves an expansion of space-time rather than movement within space-time – so no violation there either.

Other stuff…
I’m not sure that the above is an exhaustive list and I have deliberately left out other theoretical proposals such as quantum entanglement and the Alcubierre warp drive. Either of these, if real, would arguably violate relativity physics – so perhaps need to be considered with a higher level of skepticism.

25 Replies to “Astronomy Without A Telescope – Apparent Superluminal Motion”

  1. Quantum entanglement does not communicate information. The no signalling theorem tells us that violations of the Bell inequalities are due to noncausal processes. There is no underlying classical process which sets up quantum strangeness and nonlocality.
    The motion of distant galaxies with z > 1 is faster than light. In fact the CMB has a z ~ 1000, which means this region of the universe is moving away at a thousand times the speed of light. We might get alarmed at this prospect, but in fact these galaxies are not moving with a velocity in particular, but are being frame dragged by the expansion of the universe. An observer who falls into a black hole travels faster than light, but they are on a reference frame which is being dragged by the dynamics of space itself.

    The Alcubierre warp drive and related multiple connected topologies which give faster than light travel require negative energy, or that T^{00} = E < 0. The T^{ab} is the source of the spacetime curvature in the Einstein field equation. This term is ultimately determined by some quantum field, T^{00} ~ T(X,&X), and evaluated on the vacuum. If this is negative there is then no minimal quantum energy level, which is a disaster of sorts.


    1. I don’t fully understand the need for quantum entanglement as a real phenomenon. By measuring the spin of one particle, we instantly know the spin of its entangled partner. Well, sure… it’s going to be the opposite spin. Where is the need for ‘spooky action at a distance’ in this scenario?

      With respect to the redshift data, distant objects are delivering light from their distant past – so you are right that some of these objects now, today, are probably moving FTL from our position. That is the logic behind why we place the horizon of the observable universe at 46.5 billion light years from our current location – even though the universe is only 13.7 billion years old. Nonetheless, we have no reason to rule out the possibility of objects existing at 50 billion light years from our current location – they are just unknowable.

      1. Quantum entanglement of EPR pairs does get strange. The conservation of angular momentum or spin does demand that a spin 0 particle will decay into a +1/2 and a -1/2 spin particle. However, things are a bit odd. For one thing the direction of the spin is not determined by the decay. It is determined by the direction of the Stern-Gerlach apparatus you measure the spin with. So a reduction of states “here” reduces them “there.” The strangeness is also illustrated by a MRI measurement. If the spin 1/2 particle has a charge then in the presence of a magnetic field it precesses, much like a top, and there are spin flips which can occur. Now if you have an EPR pair of such spins, we can let them separate arbitrarily far apart. We have one spin particle enter a region of a magnetic field. We clearly expect that it will exhibit this precession and the spin flip with a unit of energy emitted as a photon can be detected. That is how MRI works. We change the experiment slightly. One of the particles enters the region with a magnetic field while the other enters a region far from the magnetic field with the radio frequency detector that measures the spin flips. Even though the second particle is in a region with no magnetic field it exhibits spin flips and MRI physics. Because of its entanglement with the first particle in the region with a magnetic field it exhibits the spin precession. In fact this happen independent of the distance —- they could be light years apart and so long as the entanglement persists this will happen.

        The CMB is at 47 billion light years out due to this frame dragging effect. Anything beyond the CMB is regarded as invisible, but that is only for optical radiation. The Planck probe is capable to detecting B-modes in the CMB. These are signatures of gravity waves, which were produced in the earliest moments of the universe. They are the result of gravitons decoupling from the remaining fields of nature and becoming stretched out by inflation and expansion. In the radiation dominated phase of the universe these were gravity waves which induced a certain polarization on the ~ 10000K plasma, and where these polarizations were frozen out in the end of the radiation phase about 380,000 years into the evolution of the universe. So we are attempting to measure an effect on the CMB, where the CMB is effectively a detector of much earlier physics. So in principle we can look much further back.

        The furthest back we can observe is to when the observable universe was about a string length in size. This is about 10^{-31}cm, or about 100 Planck units of length, to at most 19^{-23}cm. This is the scale at which gravitons decoupled from other fields. So if we detect B-modes we are observing indirectly physics that has a z ~ 10^{46} to 10^{36}! This means the signatures of the gravitons are stretched out to a significant portion of the horizon scale (where z = 1) we currently observe. So the region out to the CMB is just a tiny fraction of the region of the universe that is our pocket universe or nucleation bubble.


      2. “One of the particles enters the region with a magnetic field while the other enters a region far from the magnetic field with the radio frequency detector that measures the spin flips. Even though the second particle is in a region with no magnetic field it exhibits spin flips and MRI physics.”

        If the particle far away “feels” the MRI physics of the other particle, isn’t this a “signal” that can carry information?

        That is, is on Earth I turn on the magnetic field of the MRI machine, a person in Alpha Centari that observes the other particle would not notate sometning strange, so that :

        -If nothing strange happens (no MRI physics), the message is “no”
        -If something strange is measured (MRI physics measured), the message is “yes”.

      3. What you can’t control is the direction of the spin in the magnetic field. If spin is alligned with the field there is no meaured spin flip, if the spin is alligned there is a spin flip. This means you could not use the system to communicate a binary string.


      4. I don’t fully understand the need for quantum entanglement as a real phenomenon. By measuring the spin of one particle, we instantly know the spin of its entangled partner. Well, sure… it’s going to be the opposite spin. Where is the need for ‘spooky action at a distance’ in this scenario?

        I don’t fully understand what you are saying here. Are you perhaps conflating quantum entanglement as an observable phenomena, or its reality, with a need for “action at a distance”?

        As you yourself points out, quantum correlations emerging out of observing entangled quantum phenomena are mere correlations. They have more physics than classical correlations, because you can choose correlation, and delay the choice until the very end, but they are still mundane correlations at that end. There is no causal action involved, no signaling, no instantaneous “action at a distance” as per Newton. (With later generations adding the “spooky”.)

        There are experiments out there that tests entanglement and decoherence well respectively somewhat (the latter is just starting to become a matter of research). As for reality, one can see from above that there is no conflict. “Reality” is an old philosophical saw, but if one need a theory and a test I like the suggestion of Deutsch in “The Fabric of Reality” – if you kick it and it kicks back, it is real. The “Samuel Johnson” theory of reality, as it were.

        Adding to the constancy and non-agency of physical processes (“laws”) one adds specificity – (specific) action on reaction in classical mechanics, (specific) observation in quantum mechanics. This is the most what you can say on reality as opposed to, say, a non-specific dream world or a non-reacting platonic (idea) world. And it is eminently testable.

      5. You said (I don’t know how you do that ident trick): ‘They have more physics than classical correlations, because you can choose correlation, and delay the choice until the very end, but they are still mundane correlations at that end.’

        Yes, agree. I don’t think it’s useful to conflate entanglement as though it is a real observable phenomenon, although it does seem an effective way to model sub-atomic physics. This may have been Einstein’s view also (I understand he created the term ‘spooky action at a distance’).

    2. In this post:

      Was based on a paper:

      “Warp Drive: A New Approach”

      That proposes using vacuum energy originated from the casimir effect on circular extra dimensions. There seems not be disasters here from negative energy.

      If extra dimensions are indeed large as ADD and RS models (to this model to work, extra dimension must be far larger than the Planck lenght), it is really an exciting possibility for future engineers!

      By the way, vacuum energy (a.k.a. cosmological constant) doesn’t convert the entire universe (and any warp drive that exploits it) in a perpetual motion machine of the first kind (that is, that creates energy from nothing)?

  2. If sentience persists above our brief physical lives then you might not need FTL travel. A hypothetical eternal being has all the time in the universe. Humans have invented the notion as a kind of mental surrogate, its real value is not in the physical world, but in our capacity to imagine ourselves traveling at FTL. Certainly makes for good sci-fi 🙂

    The truth is that we will (probably) never be able to travel as biological entities through the vastness of the universe, however perhaps the undefined dark stuff we suspect exists could be a hint of a higher order of existence where such constraints are less stringent. 🙂 One can imagine such things with as much impunity as FTL travel, however getting there means leaving the physical universe as we know it. I guess each of us will find out the veracity of such an idea at the end of our brief lives. Or not.

    1. Not sure where you are going with this, but:
      a) a generational mission, where the crew have kids who become the crew and then have kids etc, could cover vast distances; and
      b) if you can develop an almost-light-speed drive, time dilation will allow you to cover vast distances in your lifetime, as long as you don’t mind that millions of years may have passed back on Earth since your departure.

    2. To Steve’s proposals I would add Oort cloud jumping. By using the local resources of stars Oort cloud comets we could eventually populate the entire galaxy. Hollowed out comets could be stationary or (preferably) movable.

      In fact, modulo that we need to learn the actual mining and habitation (likely including more fully close artificial biospheres, still only at some ~ 70 % material closure), I have been told that we have the necessary technology already. If you do the numbers apparently we wouldn’t have to wait for fusion reactors for energy, the amount of fissionable material in comets would suffice.

      Unless you have a specific show stopper (and some hundred meters of hydrogen stuffed ices is a good radiation shield), the fact is the reverse, we will be able to do this if we want.

  3. This is a pretty good simplification of a complex topic.
    I have made this argument before and not heard any satisfactory response yet, so here goes:
    If I wanted to set up a faster than light communicion system using phase velocity or group velocity componets of photons, it should be possible. All I would need is a detector set up to detect a phase shifted signal. The signal itself can be used as the message. The signal need not carry any information within itself. Using a codebook of pre-arranged sequences a series of phase shifted pulses could be sent in a certain order. The order of the signals would convey the message, there need not be any message within the signal. In theory, albeit this is a stretch, this could be manipulated into a kind of early warning system for laser weapons of the future which will travel at the speed of light. A detector placed somewhere between the laser attack and the receiving station could send a faster than light pulse or series of pulses allowing some fraction of a second to activate a counter-measure. I do think the defense industry would want to at least consider this possibility for Earth based assets.

    1. Thanks for the kind thoughts.

      I am operating at the limits of my understanding here, but:
      a) Saying that ‘the signal itself can be used as the message’ sounds like you are using the leading edge of a wave package to announce that ‘yoo-hoo, a wave package is coming’. That is information – I don’t know if this is really possible, and;
      b) I’m not sure if you can manipulate phase or group velocity in the vacuum of space (unless via some clever use of gravitational lensing??).

      1. Free space has no dispersion for EM radiation or photons, so there is no such manipulation. Even with gravitational lensing there is no dispersion. This is a consequence of the equivalence principle, which prevents any local change in a system from distinguishing that local region as in flat spacetime or in a gravity field.


    2. No way.
      Phase or group velocities only have meaning within an established wave.
      You first need to create that wave, the front of which (your signal) doesn’t go FTL.

      Throw a stone in a pond: the disturbance propagates outwards at signal speed, but wavelets within it ‘go faster’: they are born at the back of the wave, overtake it and die at the wavefront. Their speed cannot make any signal go faster than the disturbance reaches the opposite shore.

    3. Manu has the right of it.

      When you decompose an ideal “light on” intensity step function into Fourier components you see that you need an infinite number of them. Therefore, in reality you can’t see the leading edge go FTL for the same reason that the post already mentioned, and as Manu so vividly explains.

      The wiki links provided in the post makes a pretty good job covering all the caveats, so I hope they will provide your wished for satisfactory response. If not, perhaps you can work something better out with the wiki articles authors.

      Also, but this is now technical, fully using a predated code book isn’t actual information transfer of Shannon information at that moment. AFAIU many of code theory results doesn’t pertain to such codes for that very reason, you have to sum up all the events to make a sound analysis. So people tend to avoid those when discussing signals and codes.

      [Another similar common physics trick of “FTL communication” is turning a flashlight so that the sweep is FTL on the receiving surface. This is related to the jet illusion.]

  4. Wow, that was a fast presentation! 😀

    A little too fast, since some technical errors crept in there.

    1) The physical nature of the leading edge of a wave packet can be unclear in some cases (and the text tries to convey that), in other cases they can encode information like OAM modes. Or at least, that is AFAIU what some colleagues are trying to tell me.

    2) The inflation discussion seems a bit confused. Whether or not inflation locally exceeds speed of light, summed over cosmical distances it does (or we wouldn’t have a cosmological horizon).

    None of these inflation effects are purely theoretical or worse, “unknowable” which IMO mean “not amenable to observation or theory”. They are observable in standard cosmology (I mentioned the horizon, the early inflation expansion is observable in the CMB).

    Now for the objects embedded and dragged with these volumes they all involve expansion of spacetime, not “movement within space-time”, so there is no problem of material objects locally exceeding the speed of light. In fact, the volumes and their objects must all obey relativity. They do, and it is relativity itself which describes the objects relative movements as well as the volumes expansion.

    3) Quantum tunneling is another simple physics proposal for FTL signaling. Here the analysis becomes problematic, but the gist of it is as I understand it that it is the wave energy that one has to look at and then “the transit time” of it is a dud – it is the lifetime of the stored wave energy in the barrier that is the observable phenomena pertaining to the tunneling itself. (Cf energy levels and their lifetimes of similar constrained systems, say atoms.) Actually, the wikipedia article contains a sketch of that very argument.

    4) Even if the Alcubierre drive wouldn’t have their energy problems they would never violate relativity physics. Their ad hoc solutions for spacetime volumes are compatible with it, but they would never contain matter as they would have to be born already moving at local FTL speed. And we can’t get matter (or likely, spacetime volumes!) to exceed FTL in the first place. They are rather stupid proposals, really.

    1. Oh. I actually didn’t parse the claim on quantum entanglement correctly, since it is such a ludicrous (sorry!) claim even if it is based on a simple misunderstanding. We know that it exists and we know that it results in correlations (as described by Bell test experiments) and doesn’t involve causal signaling.

      Anyway, it didn’t make my initial list but now I have commented above instead.

  5. The article states ‘What you do is to arrange a line of light bulbs which are independently triggered. It’s easy enough to make them fire off in sequence – first 1, then 2, then 3 etc – and you can keep reducing the time delay between each one firing until you have a situation where bulb 2 fires off after bulb 1 in less time than light would need to travel the distance between bulbs 1 and 2.’

    So how do you acheive this you can’t trigger them from the previous bulb as this information will only travel to you at the speed of light as a result you can’t reduce the time delay to below the speed of light – you can’t synchronise them using electro-magnetic radiation as you are limited to the speed of light. You could have a set of clocks all set to the same time in the same place and then send one to each light location – however you then have problems with relativity as the clocks having been accelerated will all show different times.

    I thought that apparent superluminal ‘motion’ was an effect of geometry where the beam is aimed partly toward the observer.

    1. “you then have problems with relativity as the clocks having been accelerated will all show different times”
      In this thought experiment, nothing moves: the clocks would have no problems.
      Easier yet, just connect all the bulbs to a central switch with equal length wires.

  6. The main thing to understand about all this FTL business is that the ‘speed of light’ isn’t a ‘barrier’ such as the sound barrier. There is no hope of anything truly FTL because velocity > c is not impossible, it is meaningless. Speed as we understand it does not exist. Velocities don’t add. If you walk at 1 km/h in a 100 km/h train, you _won’t_ travel at 101 (nor 99) km/h. True, the difference is infinitesimal at those speeds, no more so at large fractions of c.

    The main basis for Special relativity is this: c is _constant_ for all observers, whatever their motion. Shoot a laser beam towards that Klingon vessel moving at 0.9c: all observers including Klingons, trackside cows and you will measure exactly c for that beam velocity, independently of the spacecraft motions.
    This is both an _observation_ (Michelson-Morley, check it out) and a theoretical requirement (compatibility of classical electromagnetism and mechanics).

    The only way to work around this is to discard our everyday notions of space, time, speed and replace those with spacetime, and a ‘plane shift’ (*) not unlike a rotation.

    c is the space-time constant; in natural units of length and duration its value would be 1 (which means, among other perspectives, that E=mc^2 becomes E=m: mass _is_ really energy!). Anything massless must ‘shift plane’ (*) to the maximal deviation which we perceive as c. There is no higher deviation in spacetime: c is the limit. Similarly, there is no angle ‘higher’ than 90°, if you go beyond that you fall back (another very poor similarity indeed).

    (*) very lousy analogy that just comes to mind, I can’t find better and I’m not helped by English not being my natural language 😉

    Hope this helps.

  7. I do not believe my question should be so easily dismissed.
    This is a link to a review of the now classic Princeton Univ. experiment that I had in mind:
    After reviewing the material, I will will provide a more succinct scenario for what I was thinking: A future military force is interested in protecting an important base from a laser attack, let’s call it the White House. They install a communications network around the area connected to light detectors serving as an early warning system. Let’s then say that a lossless anomalous dispersive medium is used (similar to that used in the Princeton experiment) within the cable network of the detection/warning system that is connected to a receiving station at the White House. A peak of a wave pulse (within the detection network) could be manipulated to arrive at the White House receiving station faster than a prospective laser beam sent to the White House through the air.
    There is no question that if such a transmission cable could be constructed that such a pulse could be sent faster than a laser beam. There is no question about whether or not the wave pulse could be detected as it was detected in this experiment. The question about whether or not this wave pulse constitutes a signal or not is merely a matter of semantics since the medium produces a smooth light pulse with a finite spectrum whereas traditional signals are defined as having an infinite spectrum. We know from experiments done by Gauthier that the pulse received cannot carry information within itself. That information does not propogate faster than c. But what if the pulse itself it the information? A computer at the White House could detect that a pulse is being received from the (laser attack) detector and can then have a number of nanoseconds to activate a counter-measure to an airborne laser attack before it arrives.
    The only argument that I think could be used against this is that the detector would likely have to do some computations to confirm the light pulse is an attack before sending a pulse. That would likely delay the arrival of the pulse until after the laser beam struck.

  8. There is a lot of confusion here. A linear wave will exhibit a dispersion that is itself linear. The example with water waves is a case of the Kortewig-deVreis-Boussenesque system of differential equations which are nonlinear. A light wave in vacuum is linear, and the signal it carries is strictly the speed of light. If there are nonlinear media the EM wave propagates through you can get some of these effect, but with a phase velocity that is less than the speed of light in vacuum.


    1. AFAIU, simply put, one should confuse with light speed (depending on the medium) and c (absolute light speed in a vacuum). Phase shift can be faster than the wave it travel within but not faster than c.

Comments are closed.