Measuring how cosmic structure grows

The growth of structure — often summarized by the parameter $\sigma_8$ — tells us how strongly matter clumps together over cosmic time. It is one of our sharpest probes of the dark sector: more dark energy suppresses clustering, so $\sigma_8$ encodes how dark energy and gravity shape the cosmic web.

Intriguingly, measurements of $\sigma_8$ from the recent Universe sit persistently below the value inferred from the early Universe, at the $2$–$3\sigma$ level. Together with the better-known Hubble tension, this $\sigma_8$ tension could point to unmodeled galaxy physics — or, more excitingly, to new fundamental physics.

The approach: galaxies × the CMB

My core program cross-correlates galaxy samples from the Dark Energy Spectroscopic Instrument (DESI) with maps of the cosmic microwave background (CMB). Because the CMB is lensed by all the matter between us and the surface of last scattering, galaxy × CMB-lensing cross-correlations measure the growth of structure directly — while cancelling many of the systematics that plague either probe alone.

Measuring σ₈ with DESI ELGs × Planck lensing

In Karim et al. (2025), JCAP — first author — I cross-correlated ~20 million DESI Legacy Imaging Emission-Line Galaxies with Planck CMB lensing to measure $\sigma_8$. A central result: the model used to correct for Galactic dust materially shifts the inferred cosmology, a systematic that must be controlled before claiming any tension with the early Universe.

Ongoing work

• EFT robustness with DR2 × ACT. Testing how robust Effective Field Theory modeling is for the DESI DR2 ELG sample cross-correlated with ACT CMB lensing.
• Cosmic star-formation history from ELG+LRG × CIB. Cross-correlating DESI DR2 Emission-Line and Luminous Red Galaxies with the Planck Cosmic Infrared Background to reconstruct how star formation evolved across cosmic time.
• Testing gravity with marked fields (student-led). Using marked-field × CMB-lensing cross-correlations, calibrated on COLA simulations, to constrain modified-gravity models.