Polarization Observations with the Cosmic Background Imager

TL;DR

This key agreement between the phase of the observed polarization spectrum and that predicted on the basis of the total intensity spectrum provides support for the standard model of cosmology, in which dark matter and dark energy are the dominant constituents.

Abstract

Polarization observations of the cosmic microwave background with the Cosmic Background Imager from September 2002 to May 2004 provide a significant detection of the E-mode polarization and reveal an angular power spectrum of polarized emission showing peaks and valleys that are shifted in phase by half a cycle relative to those of the total intensity spectrum. This key agreement between the phase of the observed polarization spectrum and that predicted based on the total intensity spectrum provides support for the standard model of cosmology, in which dark matter and dark energy are the dominant constituents, the geometry is close to flat, and primordial density fluctuations are predominantly adiabatic with a matter power spectrum commensurate with inflationary cosmological models.

Figures (11)

• Figure 1: The Cosmic Background Imager. For the polarization observations, the 13 90-cm Cassegrain antennae were arranged in this hexagonal close-packed configuration on the rotating, alt-az mounted platform, with six adjusted to be sensitive to right-hand circularly polarized radiation and seven to left-hand circularly polarized radiation. By correlating the signals from the antennae in pairs, 78 interferometer baselines are obtained ranging in length from 1.0 to 3.5 m.
• Figure 2: The sky coverage of the four fields imaged by the CBI in polarization. The grey-scale shows the noise-level achieved in total intensity, $I$, in the observations reported here. Three of the fields have been mapped with 36 separate pointings, whereas the fourth, $20{^{\rm h}}$, has been mapped more deeply but in only six pointings. The modulation of the sensitivity by the CBI primary beam is apparent. The approximate centers of the four fields are: $02{^{\rm h}}49{^{\rm m}}30{^{\rm s}}$, $-02^{\circ}52'30"$; $08{^{\rm h}}47{^{\rm m}}30{^{\rm s}}$, $-02^{\circ}47'30"$; $14{^{\rm h}}45{^{\rm m}}30{^{\rm s}}$, $-04^{\circ}07'30"$; $20{^{\rm h}}49{^{\rm m}}30{^{\rm s}}$, $-03^{\circ}30'00"$ (J2000 right ascension and declination).
• Figure 3: Location of CBI and DASI fields in relation to the Galaxy. The sky image is the Ka-band synchrotron map derived from WMAP first year data Bennett03fg. Galactic longitude increases to the left, with zero in the center of the image.
• Figure 4: Images of the 14$^{\rm h}$ field mapped by the CBI in Stokes parameters $I$, $Q$, and $U$ (Stokes $V$, circular polarization, is not measured and is assumed to be zero). Color is used to represent intensity, with the same scale in each Stokes parameter. In these images the contaminating effects of ground radiation and foreground emission have not been removed. The total intensity, $I$, image ( left) is dominated by CMB emission (modulated by the instrumental point-spread function); some foreground point sources are visible ( red spots). The linear polarization, $Q$ and $U$, images ( center and right) are dominated by instrumental noise and ground pickup. Ground pickup, which with our observing strategy should be the same in each pointing at the same declination, gives rise to a pattern that repeats at intervals of 3 min in right ascension.
• Figure 5: The effect of lead minus trail differencing. Here the data presented in Fig. \ref{['fig:images']} have been differenced: each visibility measurement has had the corresponding measurement on a field 9 min later in right ascension subtracted. Because the ground pickup is very similar for both measurements, ground emission cancels out in the difference. In the resulting images foreground point sources may appear positive or negative in $I$ depending on their right ascension. The $Q$ and $U$ images show that ground pickup has been removed with high accuracy.