Tentative Detection of Quasar Feedback from WMAP and SDSS\n Cross-Correlation

Suchetana Chatterjee, Shirley Ho, Jeffrey A. Newman, Arthur Kosowsky

Abstract

We perform a cross-correlation analysis of microwave data from the Wilkinson Microwave Anisotropy Probe (WMAP) and photometric quasars from the Sloan Digital Sky Survey (SDSS), testing for the Sunyaev-Zeldovich (SZ) effect from quasars. A statistically significant (2.5 $\sigma$) temperature decrement exists in the 41 GHz microwave band. A two-component fit to the cross-correlation spectrum incorporating both dust emission and SZ yields a best-fit $y$ parameter of $(7.0 \pm 3.4)\times 10^{-7}$. A similar cross-correlation analysis with the Luminous Red Galaxy (LRG) sample from SDSS gives a best-fit $y$ parameter of $(5.3 \pm 2.5)\times 10^{-7}$. We discuss the possible physical origin of these signals, which is likely a combination of SZ effects from quasars and from galaxy clusters. Both the Planck Surveyor satellite and current ground-based arcminute-resolution microwave experiments will detect this signal with a higher statistical significance.

Introduction

In modern cosmology, galaxy evolution is understood to be heavily influenced by "quasar feedback." Supermassive black holes at the centers of galaxies, when active as quasars (Active Galactic Nuclei or AGN), release enormous amounts of energy. This energy heats the surrounding gas and can potentially halt star formation, yet direct measurement of this feedback remains challenging.

The Sunyaev-Zeldovich (SZ) effect provides a unique observational window. It occurs when low-energy Cosmic Microwave Background (CMB) photons scatter off hot electrons in the gas surrounding a quasar, gaining energy and creating a characteristic spectral distortion—specifically, a temperature decrement at frequencies below 217 GHz. Because the SZ signal from an individual quasar is extremely faint, this study utilizes statistical cross-correlation to detect the aggregate signal from a large population.

Methodology

The core of this study is a cross-correlation analysis between microwave maps and large-scale galaxy surveys. By stacking microwave signals at the known positions of hundreds of thousands of quasars, the researchers effectively reduce the noise floor.

Key technical procedures included:

  • Spatial Wiener Filtering: A high-pass filter was applied to the WMAP maps to isolate small-scale SZ signals from the dominant large-scale primary CMB fluctuations.
  • Two-Component Modeling: To account for contamination from thermal dust emission in quasar host galaxies, the team performed a joint fit for the SZ $y$-parameter (tracing gas pressure) and a dust emission component.
  • Systematic Masking: Radio-loud quasars identified in the NVSS survey were masked to prevent their emission from filling in the expected SZ decrement. Additionally, standard Galactic plane masks were applied to the CMB data.

Datasets

The analysis integrated several major astronomical datasets, requiring careful consideration of sky coverage and foreground contamination.

  • WMAP 5-year maps: Specifically the Q (41 GHz), V (61 GHz), and W (94 GHz) bands.
  • SDSS Photometric Quasar Catalog: A sample of approximately 586,000 quasars.
  • SDSS Luminous Red Galaxy (LRG) Catalog: A control sample of roughly 911,000 LRGs used to trace the signal from galaxy clusters.
  • NVSS (NRAO VLA Sky Survey): A 1.4 GHz radio survey used for source masking.

Key Results

The study achieved a 2.5 $\sigma$ detection of a temperature decrement in the 41 GHz band at the positions of the quasars. The best-fit Compton $y$-parameter for the quasar sample was $(7.0 \pm 3.4) \times 10^{-7}$. For comparison, the LRG sample yielded a similar $y$-parameter of $(5.3 \pm 2.5) \times 10^{-7}$.

The observed signal is roughly an order of magnitude higher than predictions for isolated quasars but lower than expectations for massive galaxy clusters. This suggests that the detected signal originates from quasars residing in moderately massive groups or clusters, representing a combination of energy injected by the quasar itself and the ambient hot gas of the host environment.

Discussion and Future Work

A primary limitation of this study is the angular resolution of WMAP (13 to 53 arcminutes), which prevents the spatial separation of gas immediately surrounding the quasar from the larger-scale cluster gas.

Future observations from the Planck satellite and ground-based experiments like the Atacama Cosmology Telescope (ACT) and the South Pole Telescope (SPT) will provide the arcminute-scale resolution necessary to "zoom in" on these sources. Such high-resolution data will allow for a cleaner measurement of quasar feedback, potentially isolating the specific energy contribution of supermassive black holes to the intergalactic medium.

Summary updated

Abstract

We perform a cross-correlation analysis of microwave data from Wilkinson\nMicrowave Anisotropy Probe and photometric quasars from the Sloan Digital Sky\nSurvey, testing for Sunyaev-Zeldovich (SZ) effect from quasars.\n A statistically significant (2.5 $\\sigma$) temperature decrement exists in\nthe 41 GHz microwave band. A two-component fit to the cross-correlation\nspectrum incorporating both dust emission and SZ yields a best-fit $y$\nparameter of $(7.0 \\pm 3.4)\\times 10^{-7}$. A similar cross-correlation\nanalysis with the Luminous Red Galaxy sample from Sloan gives a best-fit $y$\nparameter of $(5.3 \\pm 2.5)\\times 10^{-7}$. We discuss the possible physical\norigin of these signals, which is likely a combination of SZ effects from\nquasars and from galaxy clusters. Both the Planck Surveyor satellite and\ncurrent ground-based arcminute-resolution microwave experiments will detect\nthis signal with a higher statistical significance.\n
BibTeX
@article{chatterjee2010tentative,
  author = {Suchetana Chatterjee and Shirley Ho and Jeffrey A. Newman and Arthur Kosowsky},
  title = {{Tentative Detection of Quasar Feedback from WMAP and SDSS\textbackslash\{\}n Cross-Correlation}},
  journal = {The Astrophysical Journal},
  year = {2010},
  doi = {10.1088/0004-637x/720/1/299},
}

Loading related content...

Figure