Annu. Rev. Astron. Astrophys. 1992. 30: 653-703
Copyright © 1993 by . All rights reserved

Next Contents Previous

4.4 Confusion by Galactic Dust Emission

At high frequencies the most serious source of foreground confusion is Galactic dust emission. The most detailed study of the Galactic dust emission comes from observations made with the IRAS satellite at 12, 25, 60 and 100 µm. Beichman (1987) has given an excellent review of the IRAS view of the Galaxy, including a discussion of the diffuse Galactic emission.

One of the many important discoveries of the IRAS satellite was that of the ``infrared cirrus'' radiation (Low et al 1984). Fluctuations are seen on all angular scales from ~ 2', the resolution limit of IRAS, to tens of degrees. The spectrum of the fluctuations has an approximately power-law dependance on angular scale, ~ k-1.2 (Beichman 1987), with significant power on the angular scales which have been so intensively studied in microwave background radiation observations. Over much of the frequency range of interest here, the spectrum of the Galactic dust emission is given approximately by IGde(nu) propto nu p Bnu (Tdust), where Bnu is the Planck function. The power law spectral index by which the Planck function is scaled, p, depends upon the emissivity of the dust and lies is the range 1-2. One of the critical aspects of the IRAS observations has been the difficulty of calibrating the data on the diffuse emission on an absolute scale (Rowan-Robinson 1986). This made it difficult to determine p in comparisons of the IRAS 100 µm observations with lower frequency observations, but the situation has now been resolved by the COBE results which indicate that p ~ 1.65 for a single dust components at 23.3K, and that the IRAS specific intensities are about a factor 1.67 high (Mather et al 1990, Wright et al 1991).

Masi et al (1991) have attempted to model the anisotropy of the diffuse emission at millimeter and submillimeter wavelengths based on the observations of Lubin et al at (1985) 90 GHz and the IRAS 100 µm emission, scaled to fit the COBE observations. They also included the effects of the Galactic synchrotron emission at low frequencies as observed in the 408 MHz survey of Haslam et al (1982) and the 24.5 GHz observations of Fixsen et al (1983). Their work indicates that the minimum in the temperature fluctuations due to the combined effects of the contributions from the dust and synchrotron components of Galactic emission is at numinfluct ~ 125 GHz. In an independent study (in preparation) we have calculated that the minimum in mean temperature due to these two components is at numin ~ 100 GHz. If it is assumed that the fractional fluctuation on one degree scales in the Galactic dust emission is ~ 40% - about twice that of the fractional fluctuation in the synchrotron emission - this leads to a minimum in temperature fluctuations from the combined effects of the contributions from the Galactic synchrotron and dust emission at numinfluct ~ 85 GHz. The corrseponding minima in intensity are at numin ~ 65 GHz and numinfluct ~ 55 GHz.

IRAS was not intended primarily as an instrument for accurately surveying the diffuse emission. However, significant progress has been made recently in deriving a better calibration for the observations and allowing for the contribution of zodiacal light. This has led to a very substantial improvement in the quality of the measurements of fluctuations in the diffuse components - the so-called ``Supersky Survey'' (Wheelock 1991). There is now the possibility of scaling the IRAS data to fit the COBE data in order to derive an all sky map from the IRAS data, and with the full resolution of the IRAS observations of ~ 2', with considerable precision (Hauser et al 1991).

It is clear that the Galactic foreground emission will have to be subtracted in observations of the microwave background radiation at high frequencies. The ratio of the specific intensity of the Galactic thermal emission from dust to that of the microwave

Next Contents Previous