🔭 astrophysics

TDCOSMO: XX. WFI2033--4723, the First Quadruply-Imaged Quasar Modeled with JWST Imaging

This paper presents the first lens model for a quadruply-imaged quasar derived from JWST imaging, demonstrating that advanced PSF modeling techniques significantly improve the precision of time-delay cosmography and reduce uncertainties in Hubble constant measurements compared to previous HST-based models.

Original authors: D. M. Williams, T. Treu, S. Birrer, A. J. Shajib, K. C. Wong, T. Morishita, T. Schmidt, M. Stiavelli

Published 2026-08-27

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: D. M. Williams, T. Treu, S. Birrer, A. J. Shajib, K. C. Wong, T. Morishita, T. Schmidt, M. Stiavelli

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The universe is expanding, and for decades, astronomers have been trying to measure exactly how fast. This rate of expansion, known as the Hubble constant, is a fundamental number that tells us the age and size of the cosmos. However, a stubborn problem has emerged: different methods of measuring this speed give different answers. One method looks at the light from the very early universe, while another measures the distances to nearby stars and galaxies. These two approaches do not agree, and the gap between them has only widened as measurements have become more precise. Resolving this disagreement is critical because it could mean our current understanding of physics is incomplete, requiring a revision of the standard model of cosmology.

To settle the score, scientists need a third, independent way to measure the expansion rate. One of the most promising tools for this is a phenomenon called gravitational lensing. When a massive galaxy sits directly between Earth and a distant, bright object like a quasar, its gravity bends the light from that background object. If the alignment is perfect, the single background quasar appears as four separate images arranged around the foreground galaxy. Because light travels different paths to reach us from each of these four images, the light from each image arrives at a slightly different time. By measuring the delay between these arrivals and modeling the mass of the foreground galaxy, astronomers can calculate the expansion rate of the universe without relying on the other two methods.

The challenge has always been that modeling the mass of the foreground galaxy requires incredibly sharp images. The foreground galaxy is not just a smooth blob; it has a complex structure, and its light is often mixed with the light of the background quasar. To untangle this, astronomers need to know exactly how their telescope blurs the light, a property called the point spread function. For years, the Hubble Space Telescope provided the best images for this work, but even Hubble has limits. A new generation of telescopes, led by the James Webb Space Telescope (JWST), offers a chance to see these systems with unprecedented clarity, potentially removing the uncertainties that have held back progress.

In this study, a team of researchers applied the power of the James Webb Space Telescope to a specific system known as WFI2033–4723. This is a quadruply imaged quasar, meaning the background object appears as four distinct points of light surrounding a foreground galaxy. The team used JWST to capture detailed images of this system and built a new, highly precise model of the gravitational lens. Their goal was to see if the superior resolution of the new telescope could improve the measurement of the time delays and the mass distribution of the lensing galaxy, which are the key ingredients for calculating the expansion rate of the universe.

The researchers found that the new telescope data allowed them to see details that were previously invisible. In the older Hubble images, the background quasar's host galaxy appeared as a faint, blurry smudge. With JWST, the host galaxy was clearly resolved, revealing a ring-like structure and even specific star-forming regions within it. This clarity was crucial because it allowed the team to map the mass of the foreground galaxy with much greater accuracy. A major source of error in previous models was the difficulty in accounting for smaller, nearby galaxies and satellite objects that tug on the light. In the older HST models, the mass of a small satellite galaxy near the main lens was essentially unconstrained, meaning the model could not tell how heavy it was. The new JWST data, however, allowed the team to measure the mass of this satellite directly.

By incorporating these new measurements, the team achieved a significant improvement in precision. They found that their new model reduced the uncertainty in the key time-delay measurement by 22 percent compared to the best previous models based on Hubble data. This level of precision is a record for this type of system. The results also revealed that the older models had been compensating for the missing information about the satellite galaxy by adjusting other parts of the model, such as the shape of the main galaxy's mass distribution. With the satellite now properly accounted for, the model of the main galaxy changed, leading to a slightly different calculation for the expansion rate. Specifically, the new data suggests a value for the Hubble constant that is about 3 percent higher than what was derived from the Hubble Space Telescope data for this specific system.

The study also demonstrated that the complex way the James Webb Space Telescope blurs light can be modeled with sufficient accuracy to trust the results. The team developed and tested new techniques to reconstruct the telescope's point spread function, ensuring that the sharp details seen in the images were real and not artifacts of the instrument. They showed that these new methods could handle the intricate patterns of light without introducing errors. The findings confirm that the James Webb Space Telescope is a powerful tool for cosmology, capable of refining measurements that were previously thought to be limited by the quality of available data.

While this paper focused on the technical improvements in modeling and did not calculate a final, definitive value for the Hubble constant, it lays the essential groundwork for doing so. The researchers noted that a full cosmological analysis, which would include updated data on the environment surrounding the lens, will be presented in a future paper. However, the current work proves that the next generation of lens models is attainable. By resolving the faint details of the lensed galaxies and accurately measuring the mass of nearby perturbers, the James Webb Space Telescope is opening the door to more precise measurements of the universe's expansion, potentially helping to solve one of the most pressing mysteries in modern physics.

Technical Summary: TDCOSMO XX. WFI2033–4723, the first quadruply imaged quasar modeled with JWST imaging

Problem Statement
Time-delay cosmography offers a unique, independent measurement of the Hubble constant ( $H_0$ ), which is critical for resolving the "Hubble tension" between early- and late-Universe measurements. However, the precision of these measurements is currently limited by systematic uncertainties in strong lens modeling, particularly the mass-sheet degeneracy (MSD) and errors in the point spread function (PSF). While the Hubble Space Telescope (HST) has been the standard for such modeling due to its stable PSF, ground-based seeing-limited images lack the necessary precision. The James Webb Space Telescope (JWST) offers superior resolution and sensitivity, but its complex, extended PSF presents a significant modeling challenge. This paper addresses whether JWST imaging can be modeled with sufficient accuracy to improve upon HST-based constraints for a quadruply imaged quasar, specifically WFI2033–4723.

Methodology
The authors present the first lens model for time-delay cosmography derived from JWST data, applied to the system WFI2033–4723. The analysis focuses on a direct comparison with previous HST-based models (Rusu et al. 2020) to isolate the impact of the new data and modeling techniques.

Data: The study utilizes JWST NIRCam imaging in the F115W band, which offers higher spatial resolution ( $\sim0\farcs0307$ ) than HST. Ancillary data, including spectroscopic velocity dispersions and wide-field photometry for external convergence ( $\kappa_{ext}$ ), are retained from the previous H0LiCOW XII analysis to ensure a fair comparison.
PSF Reconstruction: A critical component of the work is the development of new PSF modeling techniques to handle JWST's complex PSF. The authors compare two methods:
1. PSFr: A standard algorithm involving sub-pixel centroiding and stacking.
2. STARRED: A novel method utilizing wavelet regularization (starlets) to sparsely represent the PSF, effectively denoising the model and improving astrometric precision.
  The authors implement an iterative process where the PSF is updated based on the quasar point sources during the forward modeling of the lens.
Modeling Framework: The analysis uses the lenstronomy software package. The main deflector is modeled with a Power-law Elliptical Mass Distribution (PEMD). The quasar host galaxy is reconstructed using a Sérsic profile augmented with linear shapelet functions to capture complex extended emission. The model includes explicit treatment of nearby perturbers (satellite galaxy X and galaxies G2, G3, G7) and external shear.
Systematic Testing and Weighting: The authors explore 24 model configurations varying in PSF generation method, source complexity (shapelet order), mask size, and the inclusion of free flexion terms. These models are weighted and combined using a Bayesian approach. Weights are assigned based on:
1. Imaging Likelihood: Evaluated via the Bayesian Information Criterion (BIC) to penalize overfitting.
2. Kinematic Consistency: A Gaussian likelihood weighting based on the agreement between the model-predicted and observed stellar velocity dispersion of the main deflector.

Key Contributions

First JWST Cosmographic Lens Model: This work establishes the feasibility of using JWST imaging for high-precision time-delay cosmography, demonstrating that the telescope's complex PSF can be modeled accurately.
Advanced PSF Techniques: The implementation of the STARRED algorithm with wavelet regularization is shown to significantly reduce astrometric uncertainty compared to traditional PSF modeling techniques.
Satellite Mass Constraints: The superior resolution of JWST allows for the first robust constraints on the mass of a satellite galaxy (X) in this system, which was previously unconstrained in HST data.
Direct Comparison: By holding external convergence and other non-imaging parameters constant, the paper isolates the improvements in lens modeling precision attributable solely to JWST imaging and the new PSF methods.

Results

Precision Improvement: The JWST-based models achieve a 22% increase in precision for the Fermat potential difference ( $\Delta\tau_{BC}$ ) compared to the best HST models. The fractional uncertainty drops from $\sim3.7\%$ (HST) to $\sim2.2\%$ (JWST).
Parameter Recovery: The recovered Fermat potential difference is consistent with HST results within $1\sigma$ , but the central value is approximately 3.1% higher. If all other factors were held constant, this would shift the inferred $H_0$ for this lens from 71.6 to 73.8 km s $^{-1}$ Mpc $^{-1}$ .
Satellite Constraints: The satellite galaxy X, previously found to have a mass consistent with zero in HST models, is constrained to an Einstein radius of $\theta_{E,X} = 0.100^{+0.005}_{-0.005}$ arcsec, aligning with theoretical mass-to-light expectations.
External Shear: By properly constraining the satellite mass, the model no longer needs to "absorb" unmodeled mass into the external shear term. Consequently, the inferred external shear ( $\gamma_{ext} = 0.088$ ) is 21% lower than previous HST-based estimates, bringing it into closer agreement with observed shear estimates.
Astrometry: The JWST models achieve an overall relative astrometric uncertainty of 0.34 mas (approx. 1/100 of a NIRCam pixel), a significant improvement over HST constraints.

Significance and Claims
The paper claims that JWST imaging, when paired with advanced PSF modeling techniques like STARRED, can produce lens models with precision exceeding that of HST. This capability is essential for the next generation of time-delay cosmography, particularly for systems that are too faint or complex for HST to resolve effectively. The authors emphasize that while the results are consistent with previous HST models, the reduction in uncertainty and the ability to resolve sub-structures (like the satellite galaxy) directly address key sources of systematic error. The work substantiates the potential of JWST to expand the sample of usable lenses for $H_0$ measurements, laying the groundwork for future full cosmographic analyses that will include updated external convergence sampling. The authors explicitly state that this paper focuses on the lens modeling comparison and does not perform a final determination of $H_0$ , which will be presented in a subsequent paper.

More like this