About me

I use strong gravitational lensing to measure the expansion of the Universe and to study how massive galaxies grow. My work combines precision cosmography, deep follow-up observations, and open-source software so that large lens samples can be analyzed reproducibly.

At the University of Chicago, I lead programs spanning HST, JWST, and Rubin-era data. I co-develop community software and mentor students through projects that regularly become papers and public tools.

Research output 18 first-author, total 77 refereed publications
PI & Co-PI HST and JWST funded programs
Software lenstronomy, dolphin
Mentoring 26 mentees on 7 papers

Research highlights

DES J0408 lens system

Time-delay cosmography

The \(H_0\) tension — the \(>\)5\(\sigma\) disagreement between CMB-based and distance-ladder measurements of the Hubble constant — is one of the most pressing puzzles in modern cosmology. Time-delay cosmography offers a fully independent local-Universe route to \(H_0\). I led the analysis of DES J0408\(-\)5354, a rare double-source-plane lens, obtaining what was at the time the most precise \(H_0\) measurement from a single system (3.9%; Shajib et al. 2020). As a core member of the TDCOSMO collaboration — corresponding author on the most recent collaboration paper (TDCOSMO 2025) — I am now using JWST NIRSpec IFU data from programs I co-PI to push toward a 2% \(H_0\) measurement free of any mass-profile assumption.

RXJ1131 NIRSpec IFU data

Breaking the mass-sheet degeneracy

The dominant systematic in time-delay cosmography is the mass-sheet degeneracy — an exact symmetry of lens imaging data that leaves \(H_0\) unconstrained without external mass-profile information. In Birrer, Shajib et al. (2020), we built a framework to constrain \(H_0\) and the galaxy mass profile jointly from stellar kinematics, relaxing the standard power-law assumption at the cost of widening the \(H_0\) uncertainty from 2% to \(\sim\)8%. Spatially resolved kinematics is the key to recovering that precision. I obtained the first resolved stellar kinematics for a time-delay lens with Keck/KCWI and demonstrated a factor-of-seven improvement in constraining power over unresolved measurements (Shajib et al. 2023). Adding JWST NIRSpec IFU data for a single system has since tightened the \(H_0\) uncertainty from \(\sim\)8% to \(\sim\)5%, with seven more systems in hand.

Project Dinos logo

Project Dinos

How baryonic feedback has shaped massive elliptical galaxies over the last seven billion years is one of the open questions in galaxy formation. I founded Project Dinos — a funded HST archival program — to address this through the largest uniformly modeled sample of galaxy–galaxy lenses. Paper I delivered state-of-the-art models for \(\sim\)80 systems (Tan, Shajib et al. 2024), setting a new benchmark in sample size. Paper II placed the tightest constraints to date on the redshift evolution of dark matter halo slopes at \(z \lesssim 1\), finding broad consistency with IllustrisTNG predictions (Sheu, Shajib et al. 2025). A companion HST Schedule Gap program (500 orbits, co-PI) will extend the sample sixfold over the next two years.

Rubin LSST strong lensing forecast

The Rubin era: 100,000 strong lenses

The Vera C. Rubin Observatory's LSST is set to discover roughly 100,000 strong gravitational lenses — a hundredfold increase over the current census — transforming strong lensing into a statistical cosmological probe on par with galaxy clustering and weak lensing. I wrote the invited review charting the science cases, discovery strategies, and analysis pipeline requirements for this transition (Shajib et al. 2024). The forecasts in that paper — illustrated in the adjacent figure alongside the constraints from other Rubin probes — show that combining time-delay cosmography, lensing–dynamics, and double-source-plane systems from the LSST sample will yield one of the strongest dark energy constraints in the Rubin survey. As the former co-convener of the Rubin DESC Strong-Lensing Topical Team (2023–2025), I coordinated community preparation so the field is ready when the survey begins.

Dark energy constraints figure

Evolving dark energy

In Shajib & Frieman (2025), we introduced a physically motivated scalar-field \(w(z)\) parameterization that naturally avoids the unphysical phantom regime, and fit it to the most comprehensive cosmological dataset I have assembled. The constraints favor this scalar-field model over \(\Lambda\)CDM at 2.7\(\sigma\) — a meaningful hint that the cosmological constant may not be the final word on dark energy, and an independent data point in the context of the recent DESI findings.

Method figure for lensing calculations

Methods for the next data era

Analyzing thousands of strong lenses requires methods that are both physically rigorous and computationally tractable. I derived an analytic framework for lensing deflections with arbitrary elliptical mass profiles, filling a long-standing numerical gap and enabling lensing calculations for a far broader class of models (Shajib 2019). I also developed a joint lensing–dynamics formalism that makes it feasible to constrain complex dark-matter halo shapes without prohibitive computational cost. These methodological foundations underpin my automated modeling pipeline dolphin, which brings reproducible, uniform lens modeling to samples too large for case-by-case human analysis.

Integrated Sachs-Wolfe effect figure

Survey-scale cosmology

My first cosmological measurement detected the integrated Sachs–Wolfe effect at \(3.4\sigma\) by cross-correlating the AllWISE galaxy catalog with WMAP temperature maps (Shajib & Wright 2016). The adjacent figure shows the AllWISE galaxy overdensity projected across the full sky — the raw signal underlying that detection. The analysis demanded careful treatment of survey systematics across a wide range of angular scales — early training in the statistical rigor that large-scale structure work requires, and directly relevant to my current involvement in Rubin-scale dataset preparation.

Publications

For the full record, see my ADS profile.

Highly cited research papers

Selected invited reviews

Scientific software

I build open-source tools that keep lensing analysis reproducible and scalable. The main packages I maintain or co-develop are:

  • lenstronomy ★ 200+: the community-standard lens-modeling package and Astropy-affiliated ecosystem.
  • dolphin ★ 20+: an automated forward-modeling pipeline for galaxy-scale strong lenses.
  • LensingETC: a tool for optimizing multi-filter imaging campaigns.
  • raccoon: a JWST/NIRSpec spectral-cleaning tool.
  • squirrel: extracts spatially resolved kinematics from IFU spectroscopy by deblending lensed components.
  • Lens modeling tutorial ★ 10+: public notebooks and assignments used in AstroBridge and teaching.

Service to community

I treat community building as an integral part of the research enterprise. I co-led collaborative lensing efforts in Rubin DESC as Strong-Lensing Topical Team co-convener, mentored students from underrepresented backgrounds through AstroBridge, and previously organized public astronomy talks as coordinator of the Life Long Learning Program at the University of Chicago.

AstroBridge is a bridge program that provides research training to students with limited access to graduate-level opportunities. Our Bangladesh lensing project brought together 19 students and produced a co-authored research paper; 4 of those students have since been admitted to graduate programs in the US and Europe. The public code and notebooks are available as BDLensing.

  • Rubin DESC Strong-Lensing Topical Team: co-convener (April 2023 – March 2025), coordinating community preparation for LSST strong-lensing science.
  • AstroBridge: bridge program for students from underrepresented or resource-limited backgrounds; 4 alumni admitted to graduate programs in the US and Europe.
  • Lens modeling tutorial: reusable teaching material for classrooms and self-study.

Recorded talks

CV

Download the full CV.