Abstract
As demonstrated by Planck, SPT, and ACT, the abundance of Sunyaev-Zeldovich-detected galaxy clusters across mass and redshift is a powerful cosmological probe. Upcoming experiments such as the Simons Observatory (SO) will detect over an order of magnitude more objects than what previous experiments have found, thereby providing an unprecedented constraining potential. However, in order for this potential to be realised, the cluster detection and analysis pipelines will have to be built and understood to a much higher level of accuracy than has been demonstrated to date.
Here we discuss ongoing efforts towards the accurate modelling of tSZ cluster counts, focusing on the improvements regarding optimisation bias, covariance estimation, and foreground deprojection of [11, 13, 15], which are implemented in the publicly-available SZiFi package. Next, we briefly discuss the application of these improved cluster detection methods to Planck data. Finally, we introduce cosmocnc, a new cluster number count likelihood code that will be publicly available soon.