Cosmic Background

There is at least one source of extragalactic background that we should include in our estimator for Spitzer. It is not the well-known cosmic microwave background radiation, because its maximum brightness for Spitzer is 7 orders of magnitude fainter than the faintest interstellar medium. The important extragalactic background in the infrared is the integrated light of unresolved galaxies, which has recently been measured by the COBE satellite. The spectrum of the infrared background longward of 200 $\mu$m was measured using the FIRAS data (Puget et al. 1996, Fixsen et al. 1998), and the brightness from 1.25 to 240 $\mu$m was limited or detected in 10 broad bands using the DIRBE data (Hauser et al. 1998). The background was detected at 140-240 $\mu$m, but only an upper limit was possible at 100 $\mu$m and shorter wavelengths. For our background estimator, we can use a simple fit to the FIRAS spectrum. The DIRBE results are actually 26% higher than the FIRAS results at 240 $\mu$m. This is an area of active current research, and it cannot be resolved simply here. For now, we only need to have some estimate of the background that is reasonable accurate at the longest Spitzer wavelength, 160 $\mu$m; at all other Spitzer wavelengths, only an upper limit on the extragalactic background is known. A somewhat arbitrary sum of the simple fit to the FIRAS results from Fixsen et al. (1998) and another modified blackbody that brings the total up to match the DIRBE results at both 140 $\mu$m and 240 $\mu$m,

$$I_\nu = 1.3\times 10^{-5} \left(\frac{100}{\lambda}\right)^{0... ...(\frac{100}{\lambda}\right)^{2} B_\nu(22)\,{\rm ~MJy~sr}^{-1},$$ (3)

is a smooth curve that matches the existing observations.

A cosmic infrared background in the near-infrared has been tentatively detected (Dwek & Arendt 1998). We have not yet included the near infrared background in our model. It is significantly fainter than the zodiacal light, so this should not be a severe error. The Dwek & Arendt background brightness is about 0.01 MJy/sr at 3.5 $\mu$m, which is 5% of the zodiacal light in the ecliptic plane and up to 25% of the brightness at the ecliptic pole. As the near-infrared background gets better characterized, we will need to update the Spot background estimator. In order to include the background, however, we need to estimate what fraction of the total light of galaxies would actually be resolved into individual point sources. For the far-infrared background, this is also an important question, because current estimates show that reasonably deep 160 $\mu$m observations may resolve of order half of the background at that wavelength. But for the near-infrared observations the resolved fraction has not been evaluated, and it could be significant. Spitzer observers would not need to know the total light of all sources, but only the fraction that will be undetected into individual point sources.