1) Calibrating the Matter Power Spectrum using kSZ and weak lensing cross-correlations and 2) Increasing the dynamic range of the two-point correlation function in galaxy redshift surveys.

by Dr Anoma Ganguly (University of Arizona), Dr Aritra Kumar Gon (University of Edinburgh)

Friday 20 Feb 2026, 14:30 → 15:30 Asia/Kolkata

A 304 and on Zoom

Speaker: Anoma Ganguly (University of Arizona) 

 

Title: Calibrating the Matter Power Spectrum using kSZ and weak lensing cross-correlations

 

Abstract: Upcoming large-scale structure surveys promise to constrain the cosmological parameters with unprecedented precision. However, to realize the full potential of these datasets, we must accurately model the impact of baryonic physics on non-linear scales. Baryonic feedback processes redistribute gas relative to the underlying dark matter halo, resulting in a significant suppression of the matter power spectrum on small scales that, if unmodelled, biases cosmological inference. In this talk, I will discuss a new approach to resolving this uncertainty by cross-correlating the non-linear kinematic Sunyaev-Zeldovich (kSZ) effect with gravitational weak lensing. I will explain how this specific cross-correlation can isolate the gas distribution, allowing us to directly model the baryons and recover cosmological information from small-scales.

 

Speaker:  Aritra Kumar Gon (University of Edinburgh)

Title: Increasing the dynamic range of the two-point correlation function in galaxy redshift surveys.

Abstract: The density-split (DS) technique has been shown to outperform the traditional two-point correlation function (2PCF) in extracting cosmological information from galaxy redshift surveys. A key limitation of this approach, however, is that modelling starts from real space, whereas observations are in redshift space. The mapping between real- and redshift-space density fields is complicated due to peculiar velocities, creating a non-trivial disconnect between density bins in real and redshift space. In this work, we show that this difficulty can be circumvented by performing the density split and the associated modelling directly in redshift space.  Instead of volume weighting, as is done in the normal counts-in-cells, we evaluate the density around locations of galaxies/halos only, i.e., the PDF is number-weighted. We present a formalism to model this number-weighted redshift-space PDF. We also predict the DS bias directly in redshift space. Building on the indicator-field formalism, we develop a generalised halo-model description for DS auto-clustering in redshift space. This provides a physically motivated and computationally tractable framework that connects redshift-space observables directly to underlying cosmology, without the need for explicit real-space reconstruction.