Can Entropy Explain the Cold Temperatures of Outer Space?

[Desiree]

I was wondering why temperatures are so low out there in the vacuum of outer space? my understanding is that a thermometer works by interacting with the medium in which it's located. So when we are in a vacuum, there is nothing to interact with our thermometer to give off or absorb heat. So I wonder why temperatures are as low as -250 C even colder in outer space.

 

[Mr Noblet]

There are three types of heat transfer: conduction, convection, and radiation. The only heat transfer possible in a vacuum is radiation because of the lack of matter.

 

[Dale]

Following up on Mr Noblet's comments.

When an object has a given temperature it produces a characteristic spectrum called "black body radiation". The universe is filled with a background radiation that has the same spectrum of a black body at 2.7 K. So, a thermometer in deep space will radiate its own black body radiation and, until the thermometer cools down to 2.7 K, the amount of energy it radiates will be more than the amount of energy that it receives from the background radiation.

 

[Andy Resnick]

I think you have raised an interesting point: can a temperature be assigned to a vacuum? Or, must there be *something* in the vacuum (trace gas, photons, etc.)?

My thinking is that it is possible to assign a temperature to a "perfect" vacuum, but in order for there to exist a well-ordered temperature scale, the temperature must be identically zero. Which I think is also the same temperature for a perfect crystal (i.e. zero defects for all time), but I'm not sure about that.

 

[Integral]

Vacuum does not have a temperature, things in the vacuum have a temperature. The temperature of things in the vacuum depend upon what the object can "see". Any surface exposed to solar radiation will get quit hot, a surface exposed to deep space with get very cold. This is why most satellites have a spin, this promotes even heating of the surface. It would be quite possible for a exposed body to burn on one side and freeze on the other.

 

[gel]

I thought the answer to why it's so cold in space is the expansion of the universe is continually cooling it. The universe has expanded by a factor of 1000 since the http://en.wikipedia.org/wiki/Cosmic_microwave_background_radiation#Temperature" was created, cooling it from 3000K to 3K.

 

[Domenicaccio]

I think you have raised an interesting point: can a temperature be assigned to a vacuum?

How is temperature actually defined even for an object? I suck at thermodynamics to the point I don't even know how temperature is defined... :(

If it's defined in an experimental way, such as "temperature is what we measure when using a thermometer", then vacuum has a temperature, because the thermometer will show a result. Or at least I suppose that 0 can be said to be a result... :)

If it's defined as the macroscopical result of vibrations of particles, then vacuum has no temperature because it has no particles.

If it's defined in terms of how does the thing having temperature T interact with other things (heat transfer), then does vacuum interact? It doesn't conduct (zero thermal capacity?) and doesn't catch thermal radiation from the object, letting it cool, but what does it mean?

EDIT: I think that really vacuum itself has no temperature, but for practical purposes the temperature "out there" is just the temperature at which an object will end up if you bring it there. However it is not due to the vacuum itself, but to what irradiating sources (stars) are there close enough to heat the object. The vacuum is doing nothing.

 

[Andy Resnick]

"Temperature" is more interesting than I originally thought (and was taught). It seems to be intimately connected with entropy, rather than heat. I can't say too much more than that right now- I'm reading an interesting thermo book right now ("Rational Thermodynamics, Truesdell). Lots to think about.

Superficially, E = kT, so given an energy (or energy density), we can assign a temperature. But, I'm no longer sure if that's a definition of "temperature" or a statement regarding work. In any case, I can construct a matter-free region of space containing energy. We are used to mapping an electromagnetic field to temperature (black body radiation), but I haven't seen a simple discussion of say, *gravitational* energy (or spacetime curvature) to temperature. If someone knows one, I'd like to read it.

 

[Desiree]

Thanks guys. Very informative inputs so far.

 

[Domenicaccio]

"Temperature" is more interesting than I originally thought (and was taught). It seems to be intimately connected with entropy, rather than heat.

Uhm... food for thought. In electronics we always talk about equivalent noise temperature, which is a hint at the temperature-entropy connection. However I am not so convinced that such a connection is necessarily as strong as a near-identity...

For instance, is entropy necessarily accompanied by motion? Or are there some "degrees" of entropy which still persist even at 0k temperature (for example configurational entropies, i.e. defined as positional disorder between objects)?

 

[Andy Resnick]

Yes, definitely- entropy and temperature are tied together in circuits. Also digital electronics, in terms of information- 1 bit of information is equivalent to kT*ln(2) of entropy, IIRC.

I think that this discussion about assigning temperatures to various things is really about assigning *entropy* to things: blackbody radiation, vacuum, molecules, etc. I would be interested if anyone knows how to assign the temperature/entropy of non-blackbody electromagnetic distributions? A laser beam, for example.

My interest is in nonequilibrium systems- specifically biological processes. Too often, thermodynamics is treated as thermo*statics*. The common formulation of thermodynamics posits "equilibrium states", and moves forward from there. An alternative formulation begins instead from

\int \frac{dQ}{T} \leq 0

(if I have the sign correct) and leaves time explicit. Q is the heat flux, and T the temperature. You may recognize this as from Clausius, and is a definition of entropy. It is claimed that this approach is naturally suited to nonequlibrium systems, and processes as well, but I haven't learned how to do that yet...