One of the most interesting properties of the Universe is that its expansion appears to be speeding up rather than slowing down. Within the framework of Einstein’s theory of general relativity, cosmic acceleration is fuelled by "dark energy" which has the unusual property of possessing a large negative pressure.

This paper presented evidence based on new observational data (which had just been released in 2003) that the density of dark energy need not be a constant, as had been believed since the time of Einstein (1917), but may be dynamical and evolve with time.

“In this paper we used the available data on the luminosity distance (apparent brightness) of exploding stars called type Ia supernova which are widely believed to be `standard candles' and therefore of great use in establishing the properties of the universe."

Since dark energy appears to be the dominant form of matter-energy in the universe, constituting roughly 70% of the total energy density, its behavior and composition is of fundamental importance for all of physics and is a subject of considerable research on the part of physicists and astronomers working in cosmology and particle physics.

It describes a possible new discovery which, however, needs to be tested further, with much more precise observational data.

In this paper we used the available data on the luminosity distance (apparent brightness) of exploding stars called type Ia supernovae which are widely believed to be "standard candles" and therefore of great use in establishing the properties of the universe.

Study of supernova data tells us that the expansion of the universe is accelerating at present, i.e., objects in the Universe such as galaxies are moving away from each other more rapidly than they used to. This is not possible in a matter-dominated universe, since matter tends to attract and hence slows down the expansion of the universe. For the universe to accelerate, it must be dominated by some as yet unknown energy form—dark energy—with negative pressure.

The simplest form for such an energy was suggested by Einstein in 1917, though in a different context. Einstein’s "cosmological constant" is simply an added constant in Einstein’s equations which has the desired property of making the universe either accelerate (if the universe is infinite) or be static (if the universe is finite and closed on itself).

Initially Einstein's own preference was a static and closed universe. Much later, during the 1960s, the theoretical basis of the cosmological constant was established and it was shown that this term was linked to the vacuum expectation value of quantum fluctuations describing the quantum mechanical shimmering of empty space.

More recently, time-dependent forms for dark energy have been suggested and these include scalar fields (sometimes called quintessence field) as well as other theoretical constructs. However, the important question of whether dark energy is simply the cosmological constant, or if it evolves in some manner, has yet to be answered.

In this paper, we attempted to resolve this puzzle. We used the latest supernova data available at the time and performed a model-independent study of the data to check whether there was any evidence for the evolution of dark energy. Our results suggested that, taking into account statistical uncertainties, the current data cannot rule out evolving dark energy models.

For instance, a model which appeared to provide a good fit to the data had a super-negative pressure at present (its negative pressure was larger in magnitude than its energy density) which metamorphosed to significantly lower pressure in the past. Such models, which violate the so-called "weak energy condition"—usually considered to be sacrosanct—are called "phantoms."

In our paper we showed that, given the current data, it is not possible to choose between the cosmological constant and other, dynamical models of dark energy, including phantoms. These results have since been confirmed by several other papers and also extended using newer datasets. However, the precise nature of dark energy still remains an open question—one that can only be addressed once better quality data become available.

Although it has generally been recognized that dark energy need not be confined to the cosmological constant, there currently exists no favored or unique model for dynamical dark energy. Instead, a plethora of theoretical models and possibilities have been suggested in the recent literature to account for the current acceleration of the universe.

This led to the need for a complementary "model independent" approach which was proposed by one of us (Starobinsky, JETP Lett., 1998). As its name suggests, the "model-independent" approach determines the properties of dark energy from observational data, in a model-independent manner, i.e., without restricting oneself to any particular theoretical model.

An early application of the model-independent approach was to the first release of supernova data and was carried out by three of us (Alexei Starobinsky, Varun Sahni, and Tarun Saini) and Somak Raychaudhury (Phys Rev Lett, 2000). The present MNRAS 2004 paper presents a continuation of this line of research and analyses the much larger supernova sample published in 2003.

Our conclusions are entirely based on observational data which is steadily improving in quality as well as quantity. More recent data which appeared last year pointed to the existence of small systematic errors in the previous data which had been analyzed in our 2004 paper. Since our MNRAS (2004) paper we have continued applying model independent techniques to newer supernova data sets and other observations. Our results have appeared in JCAP, 2004, 2007, but the results remain inconclusive.

That is why the main question—whether or not dark energy evolves with time—is still open, and both possibilities—an exact cosmological constant, or slightly time-dependent dark energy, remain viable. (It should be pointed out that even the rather exotic possibility of "phantom" dark energy, which has super-negative pressure, is not ruled out by the present data.) We hope that the answer to this important question will come in the near future as data steadily improve and include, in addition to type Ia supernovae, the cosmic microwave background, the distribution of galaxies in space, etc.

We see it as leading to an understanding of the physical properties and nature of the most abundant form of matter in the present Universe—dark energy.

Firstly, the understanding of dark energy can crucially change human notions of our own place and role in the Universe, as well as of the fate of the Universe in the distant future. Secondly, if future observations prove that dark energy is not exactly constant and can transform into other, more conventional forms of energy, then one might see a discussion of practical implications of dark energy, including, in principle, the possibility of using dark energy as a new source of energy useful for humanity.