Dust Emission From Halos

From CMB-S4 wiki
Revision as of 09:57, 9 August 2017 by Bartlett (talk | contribs)
Jump to navigationJump to search

(Jim & Jean-Baptiste)

We have been exploring the potential effects of dust emission from clusters on SZ measurements and cluster detection efficiency. Dust emission from clusters, originally seen in IRAS data, was clearly detected by Planck. Planck Collaboration XXIII measured a correlation between the thermal SZ signal and the cosmic infrared background, while Planck Collaboration XLIII found a significant dust signal by stacking hundreds of clusters from the Planck SZ catalog. Given the spatial resolution of Planck (and of IRAS), we cannot separate the member galaxy contribution from any possible diffuse dust emission, the Planck and IRAS results are consistent with an origin from only the member galaxies.

De Zotti et al. (2016) proposed a model for the dust emission from cluster member galaxies in their study of the CORE mission, basing their approach on the model of Cai et al. (2013). The model predicts the dust emission as a function of cluster mass and redshift. We have normalized this model by stacking observations at 857 GHz of ~1000 Planck clusters from the PSZ2. We are using the normalized model to evaluate the impact of dust emission on SZ measurements.

We give some PRELIMINARY results in this posting for CMB-S4 in the following configuration. Observations bands include 40, 95, 150 and 220 GHz with map sensitivities, respectively, of 3.2, 0.9, 0.9 and 2.7 microK-arcmin and a 1-arcmin beam at 220 GHz. We use the Multi-Matched Filter (MMF), a filter optimized both spatially and spectrally to extract the SZ signal from clusters, to detect clusters in simulated maps. In this exercise, the simulated clusters are spherical and match exactly the filter’s spatial profile. The simulations include Galactic foregrounds generated with the PSM.

We assign SZ and dust signals, respectively, to each cluster according to the Planck SZ-mass scaling relation ([1]) and our normalized dust emission-mass scaling relation from above. We then compare the cluster catalogs extracted from the simulations with and without cluster dust emission to evaluate the impact of dust emission. Results are shown in the figures below.

Redshift distribution of clusters detected at S/N>5 without (black) and with (red) dust emission.
Ratio of the number of clusters detected at S/N>5 as a function of redshift.
Mass distribution of clusters detected at S/N>5 without (black) and with (red) dust emission. The dust emission enhances the MMF signal (in this configuration), boosting lower mass objects into the catalog relative to the case of no dust emission.


  • We find that dust can impact CMB-S4 (in the adopted configuration) cluster SZ signal measurements at the 10-20% level at M~1e14 Msol at redshifts around unity; the effect depends on both cluster mass and redshift.
  • This translates into an effect of ~20-30% on the cluster counts as a function of redshift. The effect BOOSTs the number of detected clusters because the dust emission INCREASES the SZ flux deduced by the filter (in this configuration).
  • This is an average over the simulated survey area. In reality, the effect depends on the exact location on the sky because the filter weights adjust as the Galactic foregrounds vary. This implies that the selection function in principle depends on sky position.
  • These results are PRELIMINARY and the object of further study.
  • These results are certainly both model dependent and dependent on the exact method of cluster extraction (in our case, the MMF). In this context, our results demonstrate that dust emission can have a non-negligible, and poorly understood, effect on cluster SZ measurements and detection.
  • Dust will be an even more important component of stacking studies where we attempt to measure the SZ effect from lower mass objects. In current models, SZ scales as M^(5/3), while dust scales linearly with M (and in fact may actually be shallower because low mass systems are likely to host more star formation activity). The dust signal therefore increases relative to the SZ signal as we move to lower mass systems in the stacks.