And the answer is: about 10102, mostly in the form of supermassive black holes. That’s the entropy of the observable part of the universe, at any rate. Or so you will read in this paper by Paul Frampton, Stephen D.H. Hsu, Thomas W. Kephart, and David Reeb, arxiv:0801.1847:
Standard calculations suggest that the entropy of the universe is dominated by black holes, although they comprise only a tiny fraction of its total energy. We give a physical interpretation of this result. Statistical entropy is the logarithm of the number of microstates consistent with the observed macroscopic properties of a system, hence a measure of uncertainty about its precise state. The largest uncertainty in the present and future state of the universe is due to the (unknown) internal microstates of its black holes. We also discuss the qualitative gap between the entropies of black holes and ordinary matter.
It’s easy enough to plug in the Hawking formula for black-hole entropy and add up all the black holes; but there are interesting questions concerning the connection between the entropy of matter configurations and black-hole configurations. They are explored in an earlier paper by Hsu and Reeb, “Black hole entropy, curved space and monsters,” which Steve blogged about here.
The entropy is so large.
From the section of the paper on the entropy bound for ordinary matter: “The use of a temperature T is justified because the entropy of a system of fixed size and total energy is maximized in thermal equilibrium.” However, in principle, there is no such thing as thermal equilibrium for a typical self-gravitating system. This is a strictly classical result; the instability is a well-known phenomenon called the “gravothermal catastrophe,” see section 8.2 of Binney and Tremaine. It is related to the fact that the heat capacity of a virialized self-gravitating system is negative. In fact, an isothermal gaseous self-gravitating sphere can reach higher entropy by evolving away from isothermality.
Now, this is all awfully old socks, and may have been dealt with in previous treatments of BH entropy and gravitational collapse. Further, I’m not sure that the use of a temperature is critical to this paper. But when I read abbreviated statements like the above, I begin to worry that people are giving short shrift to the thermodynamics of self-gravity, which can sometimes be counter-intuitive. Negative heat capacity is a funny thing.
More interesting to me, is that even with these highly esoteric configurations called ‘monsters’ and their tunneling implications on gwbhs, they still are some 20 orders of magnitude away from saturating the holographic bound.
Thats what I call an unresolved (and interesting) problem in physics.
If the universe continues to be homogeneous on scales beyond the observable universe then the total entropy would increase as the volume but the holographic bound only increases as the bounding surface.
So if the observable universe is 20 orders of magnitude below the holographic bound then the bound would be saturated at distance scales 20 orders of magnitude larger than the observable universe.
This is assuming that the holographic bound is not modified due to effects of expansion or dark energy.
Sean: thanks for linking to our paper!
Ben: the A^{3/4} bound is supposed to apply to weakly gravitating systems. To deduce the maximum entropy of a strongly gravitating system at fixed ADM energy you have to go through the monster analysis, and you find that (classically) there is no upper limit! In other words, the subtleties associated with negative specific heat, etc. are addressed in a fully general relativistic way in the monster analysis.
However, ordinary systems (stars, galaxies, etc.) *are* roughly governed by the A^{3/4} bound and that is why the total entropy of the universe in non-black hole systems is less than the entropy of a single galactic core black hole!
Hope that helps 🙂
Phil,
I think you may have meant 20 orders of magnitude smaller for saturation.
I also continue to speculate that dark energy is deeply connected with the holographic bound in some fundamental way.
e.
I still think it is really neat that a system that is describable the parameters of mass, charge, and spin can have /so much/ entropy.
One of the key notions here is that the entropy of a black hole reflects the number of ways it could have been assembled. This is very appealing, not least because it gives me the chance to ask the Big Brains for help with a question of black hole assembly that’s been bothering me for a while.
If an object falls (radially, say) into a black hole (non-rotating, say), then to a distant observer the object is never actually seen to cross the event horizon: classical null trajectories originating on the object can continue to reach the distant observer at any arbitrarily late time in history. But if we believe in any form of black hole radiation, then the black hole will (eventually) evaporate completely at some finite time in history as seen by the distant observer. So, to the distant observer looking at classical null trajectories coming up radially from the black hole, the “hole evaporated” news arrives _before_ the “object crossed the horizon” news. Doesn’t this mean that the full evaporation must, in an absolute sense, happen _before_ any object actually crosses the event horizon? including whatever material collapsed to create the hole initially.
If this line of reasoning is correct, then it would seem that the old question of what an unfortunate infalling observer sees at the singularity is mooted: he/she never sees inside the horizon because the horizon evaporates/disappears before he/she ever crosses it!
Clearly something funny is up. Can anyone offer any advice?
Thanks,
Paul Stankus
Paul,
My recollection is that while a distant observer “sees” the infalling observer gradually redshifted away and never “sees” him crossing the horizon (because, of course, she can’t see across the horizon), the infalling observer experiences crossing the horizon in a finite amount of his own proper time. This may be one of those problems that is easier to do in Kruskal-Szekeres coordinates.
In the case of a really massive black hole, the tidal forces at the Schwarzschild radius aren’t incredibly large, so it is possible that if you dropped your least favorite astronaut directly into a SMBH, the tidal forces across him would not be large enough for him to be disrupted, so he could actually cross the horizon in one piece. Something to try for “Jackass 2750 A.D.” if there’s a writers’ strike that year.
Hi Ben —
Yes, you can see directly in the Kruskal-Szekeres coordinate system that in a standard Schwarzschild geometry there are null trajectories originating on an infalling object outside the horizon which reach a distant observer at all later times in history, even though the object will cross the horizon after a finite amount of its own subjective proper time. Now, this doesn’t mean that the distant observer can “see” the infall along these trajectories in any meaningful way, though. Any actual photons generated by object in its local frame when it’s near the horizon will be vastly red-shifted when they get to the distant observer. In fact, the red-shift that the distant observer sees grows _exponentially_ in his/her reference frame, and so any photons that the object may emit in its own frame will quickly have their period stretched to beyond the age of the universe and so be unobservable in the distant frame. So in the case of the writer’s strike in the future,when the black hole infall is recorded the film will technically end in a freeze frame but will look like a fade to black.
The space-time of an evaporating black hole, however, is not purely Schwarzschild and the question is what difference this makes. My non-expert intuition is that the evaporating hole would be well-approximated by a series of Schwarzschild geometries with decreasing M, at least while the evaporation is slow. I would guess that the “last photon” emitted by the object just as it crosses the horizon does now make it out to the distant observer in finite time as the hole’s pull weakens — it’s probably easiest to see this by drawing light cones in Finkelstein coordinates — but I can’t speak completely authoritatively. If true, this would resolve the time-order paradox mentioned above, though it still leaves some lingering questions….
Paul
Spacetime thermodynamics is a bit stranger than ordinary thermodynamics. Thermodynamics is set up for closed systems and processes occur according to adiabatic deviations from equilibrium, or summations of these deviations. A system then approaches equilibrium, or maximal entropy. A part of this is that the heat capacity of materials or gases is positive. Hence there is a sort of effective attractor point at the minimum of a “bowl” that guides the system to a minimum at the bottom. In spacetime the heat capacity is negative. So instead take that bowl and turn it upside down. A ball bearing put on the peak will tend to roll away from what we previously called the equilibrium point.
Think of a black hole at a temperature, T ~ 1/(8pi M) (M the BH mass), equal to the cosmic background temperature. In our universe we might think of this as the CMB at 2.7K, without all those annoying galaxies and stars to muck up our “assume a spherical cow” model of the universe. Such a black hole would have the mass about equal to the moon and a radius about that of a BB. Now this black hole will exchange photons with its background. If it absorbs a photon it increases its mass. The entropy of the black hole is S ~ A/4, for A the area of the black hole, which in turn is proportional to the square of its mass. Further, the temperature of the black hole horizon will decrease. This indicates the departure from ordinary thermodynamics, where usually an increased temperature corresponds to an increased entropy. Conversely the black hole can emit a quanta of electromagnetic radiation or photon and lose mass. In this case the black hole entropy decreases and its temperature increases.
Now it is clear that in either case the black hole has put itself some tiny bit away from equal temperature with the rest of the universe. If its temperature is lowered the probability is greater that it will emit more photons than receive photons and its temperature increases in a runaway process of black hole evaporation. Similarly, if it absorbs a photon the probability is greater that it will absorb more photons than it emits and the black hole will grow, its entropy increases and temperature increases.
Spacetime thermodynamics is then one where we might well expect that any time in the past we observe will have a lower entropy. The state of our local region of the universe is one which has been driven to greater entropy than it had in the past, if gravitation is the dominant entropy source in the universe. We can only know this from photons which travel on our past light cone to us “here and now.” This past light cone is a projective space of rays, sometimes called a projective Lorentz group or space, because light rays in spacetime all have zero length. Yet we by various means can identify the distance to various sources and parameterize a spatial length to these projective rays. And we find that as we look back galaxies were more energetic, there were violent quasars, and ultimately to the CMB which is the deionization of a hot early cosmos. Beyond that domain, redshifted by a z ~ 1000 or so, things are hotter and extreme — and lower entropy. In spacetime and cosmology hotter means lower entropy. I use the language of the projective space here, for even though we can parameterize the spatial light rays into the past according to a decreasing entropy, the projective geometry remains the same — the laws of physics are ultimately blind to what we call entropy.
What about the future? A simple model of the universe is with the conformal group. This is a six dimensional space with a dilaton field, which extends relativity in certain ways. Now one can embed in this group a deSitter (dS) spacetime or a Anti-deSitter (AdS) spacetime. The deSitter spacetime is a hyperbolic universe in five dimensions which our universe under eternal inflation appears to be approximating better as time marches on. The Anti-deSitter spacetime is a bit odd, for it has two time directions in five dimensions and this results in a spacetime with periodic timelike paths. So the spacetime is usually partitioned off into a patch with only one pi-cycle. This has at conformal infinity a Minkowski spacetime. So the final “outcome” is itself a spacetime, but this can’t be predicted by data on 3-dim space in. So extra data must exist to make this conformal completeness. Well it comes with that dilaton field, for its wave equation has connections to gauge fields and particles, and this field data is involved with a duality with the AdS. This AdS/CFT (CFT = conformal field theory) duality has lots of interesting implications.
Now the dS spacetime has a cosmological event horizon, and our universe as its expands will become more diluted in its internal mass-energy content and begin to approximate a dS space. That cosmological event horizon will at some time in the stupifyingly distant future begin to radiate quanta in the same way that black hole quantum decay. This is Hawking-Gibbon radiation. The spacetime will respond to this by shifting the cosmological horizon outwards until as time “goes to infinity” the horizon is “gone.” The dS spacetime will have evolved into a Minkowski spacetime (a complete void), the same as with the AdS.
We might consider this to be maximal entropy of the universe, but it happens as time effectively goes to eternity. It is only made to appear finite in a Penrose conformal diagram, and Mark Twain’s comment, “When you fall to the bottom of the mine shaft, you ain’t going any further,” holds true. But this is a tad different from a usual thermodynamic approach to equilibrium in a finite time. Now those projective rays (light cones etc) connect to this conformal infinity by what is sometimes labelled by “scri” letters. This connects the data from the earliest universe to this conformal infinity. It further does so long null rays — the projective geometry of light rays that are insensitive to the entropy measure of their sources. So the conformal completeness of the universe is given by the initial and final states of the universe. These two are the start and end points of the grand path integral of the universe. Both of these points are voids, one a quantum void of unitarily inequivalent vacua, the other a Minkowski space. All that we see in between are just holographic projections that connect the two together. We are also woven in this gemish as well.
Lawrence B. Crowell
I was reading this and I made a writing error. I wrote
If its temperature is [lowered] the probability is greater that it will emit more photons than receive photons and its temperature increases in a runaway process of black hole evaporation.
I meant to say increased instead of lowered.
Lawrence B. Crowell
Pingback: Boltzmann’s Universe | Cosmic Variance | Discover Magazine