Anchor in the Shallows: Why DESI Data Misses the Universe's Acceleration
Analysis of baryon acoustic oscillations—density ripples—from 14 million galaxies in the DESI survey, combined with data from the Planck satellite, showed that dark energy might be weakening: the w0 parameter came out around –0.41 instead of –1. Even more intriguing, the Universe appears to be hitting the brakes: the deceleration parameter q0 shifted to positive values (~0.10). But an older survey, SDSS, which "sees" closer galaxies (redshift 0.15 vs. 0.295 for DESI), gives w0 ≈ –0.7 and a negative q0, pointing to acceleration. The riddle's answer lies in extrapolation: to gauge today’s expansion, we need data as near to us as possible; otherwise, the model is forced to guess the unknown. When supernovae at low redshift were added, the braking illusion disappeared.
The Universe expanded for almost its entire history, but about 5 billion years ago its expansion shifted into a new phase—acceleration, driven by mysterious dark energy. Back in the 20th century, Edwin Hubble showed that the Universe is expanding, and Georges Lemaître linked this to Friedmann's equations. Today, cosmological surveys, like sonars, probe baryon acoustic oscillations—frozen sound waves from the infant Universe—to reconstruct this history step by step. Two giants, SDSS and the spectroscopic DESI, gaze deep into space-time, but their accounts of the present day diverge. SDSS is confident: acceleration continues. DESI hesitates: maybe the Universe is already slowing down? A new study reconciles them, showing that the whole picture hinges on one critical detail—the nearest galaxies.
Imagine trying to figure out if a ship is accelerating by only watching its receding silhouette. You'd accurately measure its speed via Doppler shift, but the most important maneuver—right now—eludes you if there's no reference point near the hull. In cosmology, such an anchor is provided by measurements at very small redshifts (z). They give a "snapshot" of expansion in our epoch. SDSS, thanks to its galaxy sample MGS, reaches down to z≈0.15—almost the present. DESI's minimum effective z is around 0.295, billions of light-years farther away. The deceleration parameter q0, whose sign indicates acceleration (negative) or deceleration (positive), becomes elusive if there's nothing to latch onto nearby.
The authors performed an elegant experiment: they "blinded" SDSS by removing the low-redshift MGS sample. The result immediately shifted toward DESI: q0 became –0.10+0.33‑0.35, and the dark energy equation-of-state parameter w0 weakened from –0.71 to –0.38. Everything pointed to the apparent lack of acceleration being not a rebellion of physics, but an artifact of extrapolating into the unexplored z→0 region. As soon as they added Type Ia supernova data (Pantheon+), which excellently calibrate the distance scale at low z, acceleration was restored: q0 = –0.37±0.06, in excellent agreement with the ΛCDM model and pioneering measurements by Adam Riess. Including massive neutrinos did not budge the result.
What's next? DESI is already collecting data from the Bright Galaxy Survey (BGS)—a sample of bright galaxies that will slice the nearby Universe into tomographic bins and provide an effective redshift around z≈0.18, almost like SDSS. This will directly test whether the "tension" with ΛCDM vanishes. Moreover, precise knowledge of q0 and the dynamics of w0 is crucial for testing models of evolving dark energy, where the answer to "is the Universe accelerating right now?" could change the entire strategy for uncovering the nature of this substance. Upcoming missions like Euclid and the Roman Space Telescope will make such measurements routine, but for now, this old dispute teaches us: even in the era of giant surveys, the most valuable information often hides right under our noses.
🎯 If we trusted only DESI and CMB, we might conclude that the Universe shifted to deceleration about a billion years ago and then started accelerating again—but this is most likely an illusion caused by a lack of data near us.