One of the biggest puzzles in modern cosmology is the ongoing discrepancy in measurements of the Hubble constant (H₀) between local and early Universe probes, known as the “Hubble tension”. Since H₀ describes the current expansion rate of the Universe, it is a local quantity and can only be directly measured using nearby objects. In contrast, methods based on the early Universe, such as those using the cosmic microwave background (CMB), do not measure H₀ directly. Instead, they infer its value by assuming a cosmological model to extrapolate from the conditions 13 billion years ago to today. The fact that these two approaches yield conflicting values—with local distance-ladder measurements giving a higher H₀ than early-Universe methods—suggests that our standard cosmological model may be incomplete, potentially pointing to new physics.
New distance measurement with Type II supernovae
A team led by ORIGINS scientist Dr. Christian Vogl at the Max Planck Institute for Astrophysics (MPA), together with colleagues, has developed an independent method for measuring H₀ using Type II supernovae (SNe II). Unlike conventional approaches, this method does not rely on the cosmic distance ladder, thus providing a valuable, systematically independent verification of previous measurements. Their results provide a new, highly precise measurement of H₀ and enrich the debate about the expansion rate of the Universe.
Determining the Hubble constant requires accurate measurements of distances to astronomical objects at different redshifts. The most widely used technique, the cosmic distance ladder, relies on several interconnected steps: distances to nearby objects (such as Cepheid variable stars) are used to calibrate further reaching indicators such as Type Ia supernovae (SNe Ia), which then serve as standard candles to measure distances to faraway galaxies.
However, the reliance on multiple steps introduces possible systematic uncertainties, and different teams report slightly different results. A direct measurement based on known physics offers a valuable complementary approach, as it is affected by different systematics and does not depend on empirical calibrations. This is where Type II supernovae provide an exciting alternative.
Type II supernovae occur when massive, hydrogen-rich stars explode at the end of their lives. While their brightness varies depending on factors such as temperature, expansion velocity, and chemical composition, it can be accurately predicted using radiation transport models. This allows researchers to determine their intrinsic luminosity and use them as distance indicators, independent of empirical calibration methods.
Using machine learning to search for clues in supernova spectra
A critical step in this process is identifying the best-fitting model for each observed supernova. Key physical properties leave distinct imprints on the supernova spectrum: temperature shapes the overall continuum, expansion velocity sets the width of spectral lines via Doppler broadening, and chemical composition determines the strength of specific absorption and emission features. By systematically comparing observed spectra to simulated spectra from radiative transfer models, researchers can find the model that most accurately describes the supernova’s physical conditions. With such a well-matched model the intrinsic brightness—and thus the distance—can be precisely determined.
To make this process efficient, the team used a spectral emulator, an advanced machine-learning tool trained on precomputed simulations. Instead of running time-intensive radiation transport calculations for every supernova, the emulator rapidly interpolates between models, allowing for fast and accurate spectral fitting.
The research team applied their spectral modeling approach to a sample of ten Type II supernovae at redshifts between 0.01 and 0.04, using publicly available data not specifically designed for distance measurements (Fig. 1). Despite the limitations of the dataset, their method yielded reliable distances. By constructing a Hubble diagram from these measurements (Fig. 2), they obtained an independent estimate of H₀:
H₀ = 74.9 ± 1.9 km/s/Mpc
This value is consistent with most other local measurements - such as those from Cepheid-calibrated supernovae - and supports the tension with measurements from the early Universe. The achieved accuracy is comparable to the best methods and demonstrates that Type II supernovae are a promising tool for cosmology.
Publication:
C. Vogl, S. Taubenberger, G. Csörnyei, B. Leibundgut, ... , S. H. Suyu, and W. Hillebrandt, "No rungs attached: A distance-ladder free determination of the Hubble constant through type II supernova spectral modelling", 2025, submitted to A&A
Contact:
Dr Christian Vogl
Max Planck Institut for Astrophysics / ORIGINS Excellence Cluster
email: cvogl(at)mpa-garching.mpg.de
