ACT-CL J0215.4+0030 Thermodynamics

Associated publications: Kéruzoré, F. et al. (2020), A&A, 644, A93., Kéruzoré, F. et al. (2020), mm Universe @ NIKA2

This project is a case study in getting reliable inference out of a hard measurement problem - one where the signal is faint, the data are noisy, and there is significant contamination from unrelated sources.

ACT-CL J0215.4+0030 (ACTJ0215) is one of the most distant and lowest-mass clusters in the NIKA2 LPSZ survey, making it both scientifically interesting and observationally difficult: it appears as a faint, compact source in the data, and its signal is partially cancelled by bright point sources at the same sky location. Naively masking or ignoring those contaminants would bias the inferred cluster properties; the right approach is to model them jointly and marginalize over the uncertainty.

I built the end-to-end analysis pipeline for this cluster from raw NIKA2 data:

1. Data reduction: Constructed calibrated sky maps from raw time-ordered detector data, characterizing instrumental noise and signal filtering in the process. The pipeline I wrote is still in use by the NIKA2 collaboration for tSZ data reduction.
2. Forward modeling with joint contaminant modeling: Fit the cluster’s pressure profile using MCMC, modeling the observed map as the sum of tSZ signal and point-source emission. Contaminant fluxes were treated as nuisance parameters with informative priors, then marginalized over - correctly propagating their uncertainty into the final results rather than treating them as fixed.
3. Multi-wavelength combination: Combined the millimeter pressure profile with X-ray observations from XMM-Newton to reconstruct the full thermodynamic state of the gas (density, temperature, entropy), providing constraints on the cluster’s mass and tSZ survey observable \(Y_{500}\).

The analysis demonstrated that precise mass calibration is achievable even in the most challenging cases in the survey - a key validation for the cosmological program.