310 | Marc Kamionkowski on Dark Energy and Cosmic Anomalies

Episode Summary

Executive Summary: Sean Carroll and Mark Miankowski review the standard cosmological model (Lambda CDM), why it works so well, and why current “anomalies” matter. They cover dark matter, the cosmological constant, BAO, the Hubble tension, and new DESI/DES hints that dark energy may evolve with time—possibly even in a way that increases then decreases. The episode ends cautiously: intriguing signals, but not yet a discovery.

Main Topics: Lambda CDM as the standard cosmological model (Priority: 5/5): A broad explanation of how expansion, the CMB, galaxy clustering, dark matter, and dark energy fit together into a compact model with only a few parameters. Why dark matter is inferred beyond galaxy rotation curves (Priority: 5/5): Miankowski emphasizes that the strongest evidence for dark matter comes from the CMB and large-scale structure, not only spiral-galaxy rotation curves. Cosmological constant and dark energy (Priority: 5/5): They review the late-1990s supernova discovery of cosmic acceleration and discuss how a cosmological constant became part of the standard model, despite theoretical discomfort. Baryon acoustic oscillations and the CMB sound horizon (Priority: 4/5): The BAO feature is explained as a fossil imprint of early-universe sound waves, visible as a galaxy clustering bump and as a key standard ruler in cosmology. The Hubble tension (Priority: 5/5): Direct local measurements of the Hubble constant remain higher than values inferred from the CMB and BAO, and the gap has become more serious with improved JWST-based calibration checks. DESI/DES and possible evolving dark energy (Priority: 5/5): New galaxy-survey results tentatively suggest time-evolving dark energy, with a preferred fit that may imply dark energy increased and then decreased with time. Other anomalies and speculative extensions (Priority: 3/5): They briefly discuss the S8 tension, neutrino-mass constraints, and cosmic birefringence as additional signs that Lambda CDM may need refinement.

Key Arguments: Lambda CDM is extraordinarily successful because a small set of parameters fits both the CMB and galaxy surveys. The strongest evidence for dark matter comes from the CMB and large-scale structure; spiral-galaxy rotation curves are not the whole case. The cosmological constant became compelling only after late-1990s supernova data and was independently supported by CMB measurements. BAO provide a standard ruler from early-universe sound waves and tightly connect early-universe physics to late-time galaxy clustering. The Hubble tension persists because direct local distance-ladder measurements remain about 10% higher than CMB-inferred values. Late-time fixes to the Hubble tension generally fail because BAO and CMB jointly constrain the expansion history. Early dark energy was a promising Hubble-tension solution, but newer data have reduced its viability. DESI and DES now hint that dark energy may not be constant, though the signal is still tentative and model-dependent. If dark energy is truly evolving, the preferred fit appears to involve increasing energy density at some stage, which is theoretically very strange. DESI also sharpens cosmological neutrino-mass bounds, potentially disfavoring the inverted neutrino-mass hierarchy. Cosmic birefringence remains a small but interesting possible anomaly, though calibration makes it difficult to confirm.

Data Points: Hubble expansion excess: about 10% larger - SH0ES measurements of the local Hubble constant versus CMB/BAO-inferred values Dark matter to baryon mass ratio: about 5:1 - Standard cosmology requires roughly five times as much dark matter as ordinary baryonic matter CMB temperature uniformity: 1 part in 100,000 - The cosmic microwave background is nearly uniform, with tiny fluctuations seeding structure formation BAO scale: around 100 megaparsecs - The baryon acoustic oscillation bump appears in galaxy clustering at this distance scale Age of the expansion discovery: almost 100 years ago - Hubble’s discovery that galaxies are receding from one another Age of the CMB discovery: 1965 - Discovery of the cosmic microwave background as Big Bang afterglow Late-1990s acceleration discovery: 1998 - Supernova Cosmology Project and High-Z Supernova Team results implying accelerating expansion Cosmological constant size (qualitative): 0.000... with 120 zeros - Miankowski’s description of how tiny the observed cosmological constant is compared with naive expectations Cepheid-host sample: about 40–45 galaxies - Distance-ladder calibration sample for supernova Hubble measurements JWST validation sample: 16 hosts - Subset of Cepheid-host galaxies checked with JWST, supporting HST calibrations DESI galaxy survey scale: millions of galaxies - DESI maps redshifts for a huge galaxy sample to measure BAO across multiple bins DESI/DES binning: 6 or 7 distance bins - BAO analysis splits the survey into redshift-distance slices Birefringence hint: 0.3 degrees - Reported possible rotation of CMB linear polarization over cosmic time ACT birefringence significance: two-point-something sigma - Mentioned as a recent small hint in ACT data Neutrino species: 3 - Electron, muon, and tau neutrinos discussed in the context of cosmological mass bounds

Pivotal Quotes: "the evidence for the Hubble tension is now much stronger than the evidence they had for accelerated expansion in the late 90s" — Mark Miankowski: On why the Hubble tension is scientifically serious even without a known explanation "the preferred fit suggests that the dark energy density was increasing with time" — Mark Miankowski: Discussing the most striking and theoretically uncomfortable implication of recent DESI/DES fits "we live in a universe that we have been observing for centuries" — Mark Miankowski: Opening explanation of cosmology as an observational, model-building science

Implications: Cosmology is entering a high-precision anomaly phase: Lambda CDM still works, but future DESI/Rubin/Euclid/Roman data may decide whether the Hubble tension, evolving dark energy, or other hints are real breakthroughs or statistical/systematic noise.