Two ways of measuring how fast the universe is expanding disagree by so much that scientists have spent a decade arguing about it. Now, a rigorous new study says no easy fix is coming—even when dark energy flips its sign.
Astronomers have known for nearly 30 years that the cosmos is not just expanding, but speeding up. In 1998, two teams—one led by Saul Perlmutter, the other by Brian Schmidt and Adam Riess—independently discovered this strange acceleration by watching distant exploding stars. Their work earned the 2011 Nobel Prize in Physics, and the leading explanation they and others settled on is a mysterious force called dark energy.
Today, most scientists describe dark energy using a mathematical constant called Lambda, which pushes space apart. Combined with cold dark matter, this gives the standard model of the universe, known as LCDM. It has worked remarkably well, matching everything from the afterglow of the Big Bang to maps of how galaxies cluster across the sky.
But LCDM has one stubborn headache: the Hubble tension. This is a disagreement between two ways of measuring how fast the universe is expanding today. One method looks at the early universe and predicts what the expansion rate should be now. The other measures it directly, using nearby exploding stars and pulsing stars called Cepheids as cosmic mileposts. The two numbers simply will not line up. They differ by five to seven standard deviations—a gap so large that coincidence or sloppy measurement can no longer explain it.
One popular proposed fix is a model called LsCDM. It keeps most of LCDM's ingredients but adds one twist: dark energy was not always the same. Early on, it may have been negative, acting like gravity and pulling matter together. Only later—roughly when the universe was less than one-third its current age—did it flip to positive, pushing space apart instead. Earlier studies suggested this single change could ease the Hubble tension.
But three researchers—Sehjal Khandelwal, Abraão J.S. Capistrano, and Suresh Kumar—wondered whether that conclusion was solid. Their concern was not the physics of LsCDM itself, but how scientists usually measure disagreement between datasets. Most methods assume data behaves like a tidy bell curve, but real cosmological data does not always cooperate.
So the team gathered the latest measurements from multiple sources: the Planck satellite, the Atacama Cosmology Telescope, the South Pole Telescope, the Dark Energy Spectroscopic Instrument, and two supernova catalogs. They ran both LCDM and LsCDM through several rigorous statistical tests that do not assume tidy bell curves.
The results told two different stories. On one hand, the early-universe data from the cosmic microwave background and galaxy clustering agreed with each other remarkably well in both models. That foundation of cosmology is solid. But once the researchers added measurements from nearby supernovae, tensions reappeared for both LCDM and LsCDM. Neither model fully resolves the Hubble tension.
The universe is still keeping its secrets. But scientists are getting better at asking the right questions—and being honest about which answers hold up under scrutiny.