The anisotropy in the cosmic microwave background (CMB) has become a premier cosmological laboratory. Over the last decade the study of the intensity fluctuations has revolutionized our understanding of the global structure of the universe and the formation and evolution of large-scale structure. At the same time the polarization of the radiation has received increased attention, and indeed a future mission to study CMB polarization is listed as a top priority in recent national reports.

The CMB polarization arises from Thomson scattering of quadrupole intensity variations on the surface of last scattering. It can be decomposed into E ("gradient") and B ("curl") components; and if the intensity fluctuations are sourced by density perturbations, the polarization pattern has the E-mode form. Physically the polarization is tangential around large angular scale hot spots. The B-mode component arises only from vector or tensor perturbations on the surface of last scattering or from higher order effects (e.g., gravitational lensing) and is thus significantly smaller than the E-mode.

That the CMB should be (linearly) polarized is an inevitable consequence of nearly all cosmological models, relying simply on the presence of a (quadrupole) anisotropy at the surface of last scattering and gauge invariance. E-mode polarization encodes invaluable information about the early universe and our cosmological model beyond that in the well measured intensity.

The DASI, CBI, CAPMAP and now Boomerang experiments have detected the E-mode polarization: the present status on the EE power spectrum is displayed in Figure 1. In addition, the TE correlation has been detected by WMAP, DASI, and Boomerang. These detections highlight the promise of far more sensitive polarization measurements. The E-mode polarization provides increasingly tight constraints on cosmological parameters and the power spectrum of polarization anisotropies has a different dependence on cosmological parameters than that for temperature. A measurement of the polarization power spectrum therefore provides a way to break degeneracies in cosmological parameters determined from temperature anisotropy measurements alone. Of particular interest is a tighter constraint on the slope of the primordial spectrum, of prime importance in models of inflation. The correspondence between linear and angular scale is tighter for polarization than intensity, allowing improved constraints on features in the primordial power spectrum.

Figure 1: Experimental results on the EE power spectrum (the CAPMAP point is from its first partial season with a partial detector).

The great challenge for the next generation experiments is detection and characterization of the B-mode spectra. Only by exploring this territory well below the present levels of detection will we be able to understand the systematic challenges, the astrophysical contaminants and the requirements on data-analysis technology required to decode this last piece of the CMB puzzle.

Primordial B-Modes The most ambitious science goal of QUIET- the Q/U Imaging ExperimenT indeed for the field as a whole, is to study primordial B-mode fluctuations. Through our observations it will be possible to constrain or discover the signal of gravitational waves from the early universe. The passage of long wavelength gravitational waves through the plasma in the early universe generates a characteristic "curl" signature in the polarization of the CMB. If these gravity waves were produced by inflation their amplitude will directly constrain the expansion rate during that time, which is related to the energy scale of inflation. To produce a detectable signal, this scale would be near the GUT scale, providing us with our first probe of Unification Physics, well beyond the reach of accelerators.

Lensing B-modes In principle, the primordial gravity wave signature is cleanly separable from the polarization from density inhomogeneities by decomposition into E and B components. This picture is enriched by the existence of gravitational lensing by the intervening mass distribution in the universe. Our constraints on cosmological models are tight enough that we can predict the amplitude of this signal quite reliably, making its detection a strong test of our theories. And, as we move from detection to characterization, this measurement will provide unique constraints on massive neutrinos, dark matter velocity dispersion and the running of the spectral index.

Our main objective is achieving sufficient sensitivity and systematic immunity to determine the spectrum of B-mode polarization anisotropies with sub µK precision over the 50-2500. This will enable two major scientific advances: (1) the detection with more than 20 of the lensing of E-modes into B-modes from the matter distribution between the surface of last scattering and the present time; and (2) sensitivity to primordial gravity waves from inflation at a level of r = 0.009 for the tensor-to-scaler ratio, corresponding to an energy scale of inflation just below 10¹⁶ GeV. QUIET will also measure the E-mode power spectrum over an l range from 100 to 2500 in surveys far deeper than will be done with Planck, yielding much more sensitivity to unexpected cosmological effects or foreground sources.

Meeting this objective with a single telescope would require compromises in resolution and field-of-view. Accordingly we plan on moving the 7m Crawford Hill telescope (currently in use for CAPMAP) to the site, allowing the mapping of the fine-scale polarization anisotropy with a precision of 90 nK/pixel over 4x400 sqdg and 300 nK/pixel over 4x40 sqdg, chosen for low foreground contamination. This sensitivity is achieved with two cameras: a 397 element array of coherent receivers operating at 90 GHz; and a 91 element array at 40 GHz. For large angle polarization anisotropy, we will use the existing CBI platform and three 2m-class telescopes to achieve 100 nK per pixel maps of the CMB polarization over approximately 4% of the sky, with additional 397 element 90 GHz and 91 element 40 GHz receivers. Coherent detectors currently define the state-of-the-art for observations of both TT and EE. They allow excellent control of systematics and preclude the need to subtract large DC signals with extraordinary control as with other techniques. We have similarly designed our optical systems to reduce sources of spurious polarization to below the µK level. Science projections are made incorporating an optimized scanning strategy and realistic offset-removals.

The prime goal of QUIET is to fully characterize the spectrum of B modes in the polarization power spectrum, for l-values between about 40 and 2500. We will also measure the E-mode power spectrum with unprecedented accuracy. To achieve these objectives, we will need to probe both the character of foregrounds and instrumental systematic uncertainties to levels never before achieved.

Figure 2 shows the theoretically predicted E-mode and B-mode power spectra. QUIET will operate at two ranges of angular scales; the figures show the QUIET noise levels together with those from WMAP and Planck. QUIET small scale operations (which will use the 7m Lucent Technologies Crawford Hill Telescope to be described later) have a beam size well matched to the scale at which the lensing B-mode signature is expected to peak. Similarly, QUIET large scale observations have a field-of-view well matched to the scale at which primordial B-modes are predicted to peak. Since the spectrum of lensing fluctuations overlaps in l with the primordial gravity wave spectrum, measurements at both angular scales are necessary to unambiguously detect gravity waves. QUIET will make sensitive polarization maps at both arcminute and degree scales, providing the leverage to distinguish the source of detected B-modes.

Figure 2: The plots show the predicted power spectra together with expected noise variances from QUIET as well as WMAP and Planck. On the right (7m telescope), the concordance model power spectra for EE and BB from lensing are shown. On the left (2m telescopes), the same EE and BB spectra are shown as well as the BB spectrum with T/S = 0.18. At a given l, modes are detected with high significance when the noise spectrum is below the power spectrum.

The proposed research leads to benefits beyond scientific discoveries. Collaboration members are eager to continue strong programs of hands-on research for students- we've mentored more than sixty in laboratory research in just the last three years. We will continue to send university students to JPL for invaluable technical training and hundreds more will be informed of our science during regularly scheduled class time. Our work will contribute to increased public understanding of this most fascinating area of the physical sciences: we estimate that over the life of the project, some 100 more undergraduates will have contributed to the science and over forty public lectures will be delivered by QUIET scientists who will also help create radio and TV shows about science. The basic packaging technology developed for the compact cryogenic QUIET arrays finds application in a number of fields beyond the CMB, including arrays for use in atmospheric remote sensing, thermal mapping of extrasolar planets, and telecommunications.