 |
| This figure shows a slice of the universe mapped by the DR1 data, showing the four major galaxy samples. See the figure in the paper for more details. Credit: Claire Lamman |
The Dark Energy Spectroscopic Instrument (DESI) collaboration published a new analysis of dark energy on 19 March 2025.
DESI created the largest three-dimensional map of the universe ever made, using data from nearly 15 million galaxies and quasars collected in the first three years of observations. This map covers 11 billion years of cosmic history.
Dark energy has long been considered a constant (called the “cosmological constant” in the standard Lambda CDM model), that is, a force that does not change over time and that pushes the universe to expand faster and faster. However, the new DESI data, combined with other observations (such as those of the cosmic microwave background, supernovae and weak gravitational lensing), suggest that dark energy may not be constant. Instead, it appears that its influence may change over time, perhaps weakening. This is an important clue, because if it were true, the standard model of cosmology (Lambda CDM) would not be able to explain everything and may need to be revised.
DESI researchers looked at a feature called baryon acoustic oscillations (BAO), which acts as a “cosmic ruler.” By measuring how matter is distributed in the universe at different times in its history, they can work out how much dark energy has influenced the expansion. The new data show that this influence may vary, and the evidence for “evolving” dark energy is stronger than the results from DESI’s first year.
However, the scientists stress that more data is needed to be sure, and they are continuing to run tests to rule out errors or unknown effects.
In short, DESI is finding increasingly compelling evidence that dark energy may not be constant, but changing over time, opening the door to new discoveries about the nature of the universe.
How reliable is the hint that dark energy is evolving?
To evaluate its reliability, we must consider several aspects:
- DESI data is robust and precise: DESI used a huge amount of data (15 million galaxies and quasars) and advanced techniques such as BAO, which are considered very reliable in cosmology. Compared to the first-year results, this new analysis has more than double the data, which increases the precision and confidence in the results. The researchers also ran multiple tests to ensure that the results were not influenced by errors or unexpected effects.
- Taken alone, the DESI data are compatible with the Lambda CDM model, which assumes constant dark energy. However, when combined with other observations (cosmic microwave background, supernovae, gravitational lensing), discrepancies emerge. This suggests that a model with variable dark energy might better fit the overall data.
- Scientists speak of “clues” and “preference” for evolving dark energy, but not definitive proof. In physics, a discovery is considered certain only when it reaches a statistical significance level of 5 sigma (a probability of error of less than 0.00006%). The DESI results, combined with other data, arrive at a level between 2.8 and 4.2 sigma, depending on the combinations used. This means that there is a low probability (about 0.3% at 3 sigma) that it is a coincidence, but it is still not enough to declare an official discovery. With more data, this significance could increase or decrease.
- This is not the first time that a variable dark energy has been hypothesized. The data from the first year of DESI (2024) also suggested something similar, and now the evidence is getting stronger. However, in the past some 3 sigma signals in physics have disappeared with more data, so caution is needed.
- DESI is a five-year project, and we are only in the third year. In addition, other future experiments (such as those with more powerful telescopes) will provide complementary data. If the next results continue to support this trend, the idea will become more solid.
In conclusion, the suggestion that dark energy is evolving is very interesting and based on solid data, but it is not yet definitive. The evidence is stronger than in the past and DESI's work is serious and well monitored. However, we are not yet at the point where we can say with certainty that the Lambda CDM model is wrong or that dark energy really changes. It is a promising hypothesis, which could revolutionize our understanding of the universe, but it still needs time and further confirmation. For now, it is an idea that scientists take seriously and that deserves to be explored further.
--------------------------------
You can consult the DESI Key Publications (numerous papers delving into the technical details) at this link
DESI DR2 Results: March 19 Guide
