|Annu. Rev. Astron. Astrophys. 2001. 39:
Copyright © 2001 by . All rights reserved
3.8. Integrated Light from Extragalactic Source Counts
Galaxy number counts provide important constraints on the background light. The cumulative brightness of galaxies is a strict lower limit on the background, and the number-magnitude relation provides important information on the nature and evolution of the sources contributing to the background. Conversely, the background light provides important constraints on galaxy number counts, providing an important measure of completeness and limits on the possible existence of a truly diffuse background component.
A necessary condition for convergence of the integrated galactic light is that the logarithmic slope of the differential galaxy count per magnitude interval, dlog N / dm, should drop below a value of 0.4 at faint magnitudes. However, convergence does not ensure measurement of the total background, both because a significant fraction of the flux from galaxies can come from low-surface brightness regions missed in standard galaxy aperture photometry and because there may be truly diffuse sources of the background. The overlapping wings of resolved galaxies can create a relatively smooth unresolved background, which can only be detected by absolute surface photometry. We therefore consider integrated galaxy light as a lower limit to the background, even when the integrated light has converged. Background limits from galaxy counts are listed in Table 3.
|(µm)||I (nW m-2 sr-1)||Comments||Reference|
|0.1595||>2.9+0.6-0.4||HST/STIS||Gardner et al. 2000|
|0.1595||<3.9+1.1-0.8||HST/STIS||Gardner et al. 2000|
|0.2||0.6||FOCA||Milliard et al. 1992|
|0.2365||3.6+0.7-0.5||HST/STIS||Gardner et al. 2000|
|0.36||2.9+0.6-0.4||See Reference||Madua & Pozzetti 2000|
|0.45||4.6+0.7-0.5||See Reference||Madau & Pozzetti 2000|
|0.67||6.7+1.3-0.9||See Reference||Madau & Pozzetti 2000|
|0.81||8.0+1.6-0.9||See Reference||Madau & Pozzetti 2000|
|1.1||9.7+3.0-1.9||HST/NICMOS||Madau & Pozzetti 2000|
|1.6||9.0+2.6-1.7||HST/NICMOS||Madau & Pozzetti 2000|
|2.2||7.9+2.0-1.2||HST/NICMOS||Madau & Pozzetti 2000|
|7||1.7 ± 0.5||ISO/ISOCAM||Altieri et al. 1999|
|12||0.50 ± 0.15||ISO/ISOCAM||Clements et al. 1999|
|15||3.3 ± 1.3||ISO/ISOCAM||Altieri et al. 1999|
|25||0.02||IRAS||Hacking & Soifer 1991|
|60||0.4||IRAS||Hacking & Soifer 1991|
|90||1.0||ISO/ISOPHOT||Matsuhara et al. 2000|
|90||~ 0.9||ISO/ISOPHOT||Juvela et al. 2000|
|95||0.5||ISO/ISOPHOT||Kawara et al. 1998|
|100||0.2||IRAS||Hacking & Soifer 1991|
|150||~ 1.0||ISO/ISOPHOT||Juvela et al. 2000|
|170||0.35||ISO/ISOPHOT||Kawara et al. 1998|
|170||0.9||ISO/ISOPHOT||Matsuhara et al. 2000|
|175||1.75||ISO/ISOPHOT||Puget et al. 1999|
|180||~ 1.2||ISO/ISOPHOT||Juvela et al. 2000|
|850||0.11||SCUBA||Barger et al. 1999a|
|850||0.5 ± 0.2||SCUBA||Blain et al. 1999b|
|a Bold face indicates counts to a depth where the light has converged. Inequalities indicate limits on the integrated galaxy light, not the EBL. All entries are regarded as lower limits to the EBL. See text for abbreviations.|
Madau & Pozzetti (2000) summarized deep galaxy counts in the U, B, V, I, J, H, and K bandpasses compiled from the Hubble Space Telescope (HST) northern and southern deep fields (HDF-N and -S), supplemented with shallower ground-based observations in a variety of fields. In all of these spectral bands, the logarithmic slope of the differential number counts has dropped below 0.4 at the depth of the HDF-S survey (mAB ranging from 25.5 at K to 30.5 at V), indicating that the light has largely converged. [The AB magnitude system is defined by AB = -2.5 log F - 48.6, where F is in units of ergs cm-2 s-1 Hz-1. The AB scale is chosen so that AB at 0.55 µm corresponds very nearly to the Johnson V magnitude (Oke 1974).] The integrated light from these counts is presented in Table 3.
Gardner et al. (2000) extended the deep galaxy counts in the HDF-N and HDF-S fields to shorter UV wavelengths using the Space Telescope Imaging Spectrograph (STIS). At near-UV (NUV) (0.2365 µm) and far-UV (FUV) (0.1595 µm) wavelengths, these counts cover the brightness ranges of 23-29 and 23-30 AB magnitudes, respectively. The counts in both bands show a shallow or flat slope at the faint magnitudes sampled in these surveys. To obtain estimates of the total galaxy light at these wavelengths, they combined their data with the counts of galaxies brighter than 20.75 AB magnitude at 0.2000 µm obtained by Milliard et al. (1992) using the FOCA balloon-borne telescope.
The IRAS Faint Source Survey reached sensitivity limits of ~ 200 mJy at 12, 25, and 60 µm, and about 1 Jy at 100 µm. Hacking & Houck (1987) analyzed the extensive IRAS survey and calibration data near the north ecliptic pole, reaching a depth of ~ 50 mJy at 60 µm. Gregorich et al. (1995) analyzed a subset of IRAS pointed observations, also reaching 60 µm sensitivity levels of 50-100 mJy. They found the source density to be about twice that reported by Hacking & Houck. However, Bertin et al. (1997) also analyzed the faint IRAS 60 µm counts to the 100 mJy level, choosing fields with very low cirrus, and found results inconsistent with those of Gregorich et al. and close to those of Hacking & Houck. They suggested that the counts in the areas studied by Gregorich et al. may have been elevated by false detections due to cirrus. We have integrated the infrared light from IRAS galaxy counts as summarized by Hacking & Soifer (1991) at 25, 60, and 100 µm to the observational sensitivity limits (Table 3). The integrated light from the IRAS galaxies is far from converging and is therefore a conservative lower limit on the CIB.
The ISOCAM and ISOPHOT instruments on the ISO satellite were used for many surveys (Lemke et al. 2000). Altieri et al. (1999) carried out the deepest ISOCAM survey at 7 and 15 µm using the lensing cluster A2390, reaching limiting fluxes down to ~ 30 µJy. They combined the results of this survey with counts obtained in the HDF (Aussel et al. 1999, Taniguchi et al. 1997, Elbaz et al. 1999) to find the integrated intensity at 7 and 15 µm (Table 3). Clements et al. (1999) conducted a deep 12 µm survey in four high Galactic latitude fields with the ISOCAM, reaching a sensitivity of 0.5 mJy.
Puget et al. (1999) carried out an ISOPHOT survey at 175 µm, reaching a flux limit of ~ 120 mJy in a 0.25 deg2 area called the Marano 1 field. To obtain the integrated galaxy light, Puget et al. augmented their data with ISOPHOT 170 µm number counts from a survey of the Lockman Hole (Kawara et al. 1998) and with counts obtained by the ISOPHOT 170 µm Serendipity Survey (Stickel et al. 2000). Puget et al. found the integrated light from sources brighter than 120 mJy to account for approximately 10% of the detected CIB.
Matsuhara et al. (2000) reported the number counts and integrated light obtained with an ISOPHOT 90 and 170 µm survey of the Lockman Hole, with flux limits of 70 and 100 mJy at the respective wavelengths. Juvela et al. (2000) surveyed several fields with the ISOPHOT at wavelengths between 90 and 180 µm. For sources brighter than 100 mJy detected at more than one wavelength, they found integrated surface brightnesses at 90, 150, and 180 µm similar to those of Matsuhara et al. and Puget et al. (Table 3).
At submillimeter wavelengths, the rapid cosmological dimming of sources with increasing redshift is nearly compensated by the strongly negative K-correction for dusty sources. That is, because the intrinsic source spectrum is brighter at wavelengths shorter than the observed submillimeter wavelength, one samples intrinsically brighter parts of the source spectrum for a large range of redshifts. In principle, this makes dusty sources detectable to very high redshift (Blain & Longair 1993). Deep source counts at 850 µm have been obtained with the Submillimeter Common User Bolometer Array (SCUBA) on the James Clerk Maxwell Telescope. A population of about 20 distant galaxies has been detected at 850 µm down to a 3 limiting flux level of ~ 2 mJy (Barger et al. 1999a, and references therein). Blain et al. (1999b) extended the 850 µm counts down to a flux level of 0.5 mJy in surveys conducted through lensing clusters, accounting for nearly all the measured CIB (Table 3).