Observations of the CMB
In 1992, the NASA COBE satellite detected small (
T/T ~ 10-5) temperature irregularities in the microwave background
radiation temperature on angular scale 
(Smoot
et al. 1992). The COBE results provided the first convincing detection
of primordial temperature anisotropies, which had been searched for
earnestly by experimenters since the discovery of the microwave
background by Penzias & Wilson 1965. COBE has provided low
resolution maps of the sky at three frequencies, 30, 53 and
90 GHz and with relatively poor sensitivity (T/T ~ 10-5 at 53 GHz) after four years of
observation. The COBE
measurements are consistent with ground based measurements from
Tenerife at frequencies 10, 15 and 32.5 GHz (Watson et al.1994)
and with the balloon-borne Far-infrared Survey (Ganga et al.1993). Thus
at large angular scales ( 
5°) there is strong
evidence that the temperature irregularities are independent of
frequency over the range 10--170 GHz, as expected if they are of
primordial rather than of Galactic origin.
The COBE discovery has stimulated a concerted effort to detect
primordial anisotropies on smaller angular scales ~ 1°
(e.g. Fischer et al. 1992,
Gaier et al. 1992;
Schuster et al. 1993,
Meinhold et al. 1993, Wollack et al. 1993,
Cheng et al. 1994,
de Bernardis et al. 1994,
Gundersen et al. 1995,
Ruhl et al. 1995,
Tanaka et al. 1996,
Netterfield et al. 1997,
Platt et al. 1997,
Tucker et al. 1997; for recent compilations see
Lineweaver et al. 1996,
Rocha & Hancock 1996).
Figure 1.1 shows a recent compilation of CMB experiments (top) and
the region in frequancy and multipole space (bottom) covered by
these experiments (taken from
M. Tegmark's web page).
 |
|
Figure 1.1 (Top): CMB Experiments as of April 1999.
(Bottom): Angular and frequency sensitivity of various experiments.
For future multichannel experiments,
the crude boxes have been replaced by polygons, the highest l-value at each channel being that at which its effective noise level
(see Knox 1995 or Tegmark & Efstathiou
1996) exp(^2 l^2)/w equals that of Qflat=20 microkelvin CMB fluctuations. Likewise, the colored regions show where fluctuations in various foregrounds are expected
to exceed those of Qflat=20 microkelvin CMB fluctuations in the cleanest 20% of the sky. The foregrounds are dust (red), point-sources (green), synchrotron (magenta)
and free-free emission (cyan), and the estimates are based on Tegmark & Efstathiou 1996.
These
experiments utilize either radio frequency techniques
at dry, high altitude, sites such as the South Pole, or balloon-borne
bolometers sensitive at frequencies 150 GHz. Almost all of these
experiments report detections of temperature anisotropies
with an amplitude T/T ~ 1 - 4 x 10-5 , but the limited
sensitivity and frequency coverage of any single experiment, particularly
those at radio frequencies, precludes any firm conclusion that the
anisotropies are primordial.
To meet the science goals of this proposal, an experiment with high
angular resolution ~ 10', and
high sensitivity
T/T ~ 10 -6 is required that maps a large
fraction of the sky over a wide frequency range. As we will show in
this proposal, the latter requirements are essential to measure
the statistical properties of the CMB fluctuations to high accuracy
and to monitor and remove Galactic and extragalactic
foregrounds. These requirements are
formidable, and can only be met by a satellite experiment.
It is important to emphasize that balloon-borne experiments planned
for the next decade cannot achieve the comprehensive frequency and sky
coverage of PLANCK. For example, the TopHat experiment of
Kowitt
et al. 1995 (frequency range 150-630 GHz, angular resolution
of 0.5°) will be mounted on top of a balloon to reduce ground and
atmospheric emission, and is expected to measure temperature differences at
40 points on the sky in a single long duration flight of two weeks. The area
covered by TopHat per flight is only one four-thousandth of the area
of sky covered by PLANCK and would be observed with considerably lower
resolution. Furthermore, it is exeedingly difficult to design a balloon-borne
experiment free of spurious atmospheric and ground emissions to an accuracy
of T/T 
1 x 10-5, which is an order of magnitude higher than the
sensitivity required for the sientific goals of this proposal.
Similar considerations apply to ground based experiments, including
interferometric experiments such as the Cambridge Anisotropy Telescope
(O'Sullivan et al. 1995). A satellite experiment is the only
feasible way of simultaneously achieving a large sky coverage at high
angular resolution and sensitivity over a wide frequency range, while
eliminating atmospheric and sidelobe contamination. This has been widely
recognized by the experimental community, both in Europe and the USA,
with proposals for CMB anisotropy satellites under serious consideration by
ESA and NASA.
How different will the CMB sky look to PLANCK compared to COBE? The maps in Figure 1.2 show simulations of the CMB anisotropies expected in an inflationary cold dark matter model. The upper figure is constructed at the angular resolution and signal-to-noise of the second year COBE maps (Bennett et al. 1994). The lower figure shows how the sky would appear at the much higher angular and sensitivity of PLANCK. The individual hot and cold spots seen in the PLANCK simulation have physical sizes comparable to those of clusters and superclusters of galaxies observed in the present Universe. Evidently, we expect to see a wealth of fine-scale structure in the microwave sky. The primary goal of PLANCK is to map these structures with high precesion, free of foreground contamination, so enabling us to address the scientific questions described in the rest of this proposal.
Figure 1.2:
Simulated maps of the
background radiation anisotropies expected in an
0=1 CDM model.
The upper picture shows a simulated COBE map with a resolution of
FWHM=7°
together with random noise as in the COBE
two-year data. The lower panel shows the same realization of the sky
at the much higher angular resolution and signal-to-noise of
PLANCK.