We wish it was all that simple...

Until quite recently we had a nice, closed, cozy picture of the Universe which was good from the moments very close to the Big Bang up to the present times, and explained everything (all right, almost everything) that we knew then about the Universe. This cosmological "standard model" was based on the assumption that the Universe first endured a period of accelerated expansion (called inflation), then it passed into a deccelerated phase that lasts until now. Let's just remind what this simple picture was.

The form of matter responsible for the inflationary phase is a smooth scalar field , whose potential energy fuels the expansion rate (see figure on the left). The field (represented as the red ball) rolls very slowly down the potential, so its energy density is nearly constant. But in Einstein's General Relativity, if the energy density is constant (or almost constant), then the cosmological pressure must be negative and space must inflate like a baloon, with the distances between points increasing exponentially (or almost exponentially.)

When inflation ends, the scalar field reaches the bottom of the potential and dissipates away, leaving the Universe filled with the stuff it is made up today -- radiation, (i.e., photons, neutrinos and other light particles), which has positive pressure equal to a third of the energy density, and heavy particles (such as protons, neutrons etc.), which have zero pressure. In either case, Einstein's theory tells us that the expansion, for positive or zero pressure, must be deccelerated.

So, the Universe should be today in a deccelerating phase, if it is dominated by this "normal" kind of matter. But is it, really?

In 1998, two groups of astronomers completed their observations of a large number of exploding stars called Supernovas type IA. These stars have the curious property that the light that they emmit follows a very simple pattern. The two groups of astronomers then managed to put all the light curves from those stars together, and tried to infer with what velocity those stars were moving away from us, as a function of their distance to us. The distance from us to the star is calculated using the luminosity of the star -- since the amount of light that you can detect from a light source falls with the inverse squared of the distance to the source. The velocity of the source with respect to us can be inferred from the redshift of the light from that source. As you know, sound changes pitch as the source moves towards us or away from us. Light is not different: if a source moves away from us, its light will shift to the red (cold) end of the spectrum; if it moves towards us, light will shift to the blue (hot) end of the spectrum.

When the two teams finished their work and compiled their data, the results were mind-boggling: if their observations are correct (and if the treatment of the data is not corrupting the information, which is still a matter of some debate), the Universe is now expanding with an accelerating rate!

Now, this is quite strange: the matter that we can see (photons, protons, electrons, etc.) patently does not cause the Universe to expand like that. So, whatever this matter that is causing the accelerated expansion is, it must be something we can't see. Could it be dark matter, those invisible dark and heavy particles that are needed to stabilize galaxies such as our own Milky Way? But no, it cannot be dark matter, since it also has zero pressure, so the type of expansion it causes is also deccelerated. What we need, really, is some constant (or nearly constant) energy density which, like in the case of inflation with a scalar field, has negative pressure and causes accelerated expansion.

Dark energy (also known as "quintessence" ) is what we call this misterious type of matter with negative pressure that is making the Universe expand faster and faster. Nobody have any good idea so far as to what dark energy might be: we just know for sure what it is not: it is not photons, protons, electrons or any other form of matter seen in stars, planets, bricks or supermodel's brains; and it is certainly not dark matter either.

Well, there is not much left. One of the things we left out is a scalar field, like the one responsible for inflation; the other thing we left out is a type of matter that Einstein thought about in 1917, as he first worked out some cosmological solutions to his newly discovered theory of General Relativity -- the "cosmological constant".

What is dark energy?

1. The simplest candidate for dark energy is the cosmological constant. The cosmological constant is just that -- a constant, with a constant energy density and a constant pressure. As we discussed in the case of inflation, constant energy density means negative pressure, which means accelerated expansion. So this looks easy, just take some constant, adjust it to the values of the acceleration that we are seeing today, and end of story.

However, when one tries to implement this solution in practice, one finds that this "constant" has to be very finely adjusted. Just to have an idea of how fine this adjustment has to be, imagine that you have to paint a layer of veneer on a sphere. If this sphere were the size of the Earth, the thickness of the layer of veneer would have to be smaller than the size of an atomic nucleus! Evidently, not everybody agrees that this is a good solution.

2. The other natural candidate is a scalar field: if one can lead to inflation in the beginning of the Universe, another could lead to the accelerated phase we are experiencing now. Not as simple as the cosmological constant, but maybe this time we will not need that "fine tuning".

Not really. First, the types of scalar field potentials that can do the job of accelerating the Universe just NOW, and not some time earlier in the past, are quite strange. And second, when we actually try to fit the models with the observations, we again have to do some sort of "fine tuning" -- less than before, but still too much.

The bottom line is: there are no good candidates so far for dark energy. More likely than not, the models we have right now (which involve severe fine tunings) will turn out to be false. We need better models, which really means that we theorists don't have a clue.

Well, let's at least try to make a living with what we have -- a finely-tuned cosmological constant or some type of scalar field with a finely-tuned potential. Can we test these models against the observations?

The answer is: in principle yes; in practice, maybe. In theory, of course, each model is different than the other. But the practical problem is that the impact of the dark energy component is very small on large-scale structure (i.e., stars, galaxies, clusters of galaxies, etc.) The main impact of dark energy is on the expansion rate. Moreover, the influence of dark energy is such that it is growing now, but it was small in the past -- so, if we look too far back in the past, we will have no clue to its existence. But on the other hand, if we only look at objects now, we will be looking too near to ourselves, where it is very hard to figure out what is due to the expansion of the Universe and what is due to the peculiar motions of the galaxies.

Therefore there is a balance between near and far objects that astronomers need to observe in order to test dark energy. Nevertheless, there is still much that can be done (and is being done) to figure out what causes the Universe to accelerate.