In Figure 19, we plot EBL detections to date, together with the integrated light in detectable sources (lower limits to the EBL) in units of I between 0.1 and 1000µm. (2) The DIRBE and FIRAS detections at > 100µm and the lower limit from IRAS detected galaxies at 10-100µm indicate that energy is contained in the far infrared portion of the spectrum. Given that light from stellar nucleosynthesis is emitted at wavelengths 0.1 - 10µm, Figure 19 emphasizes the fact that 30% or more of the light from stellar nucleosynthesis has been redistributed into the wavelength range 10 - 1000µm by dust absorption and re-radiation and, to a lesser degree, by cosmological redshifting. Realistic estimates of the total energy from stellar nucleosynthesis must therefore be based on the bolometric EBL from the UV to IR. In lieu of accurate measurements in the mid-IR range, realistic models of dust obscuration and the dust re-emission spectrum (dust temperature) are needed. To discuss the optical EBL in the context of star formation, we must therefore first estimate the bolometric EBL based on the optical EBL detections presented here and current measurements in the far-IR. We do so in the following section.
Figure 19. EBL detections, limits, and models as a function of wavelength. The filed circles show the EBL detections with 2 error bars and lower limit symbols as defined in Figure 8. Also plotted are lower limits from Armand et al. (1994) at 2000Å, the HDF (Williams et al. 1996), Gardner et al. (1997) at 2.2µm, IRAS (Hacking & Soifer 1991) at 10-100µm, and Blain et al. (1999) at 450 and 850 µm. These lower limits are based on the integrated flux in detected sources at each wavelength. Upper limits marked in bold are from Hurwitz, Bowyer, & Martin (1991) at 1600Å and DIRBE (Hauser et al. 1998). The open circles indicate DIRBE detections. The bold line at 125-1000µm shows the FIRAS detection (Fixsen et al. 1998). Detections in the near-IR are from Wright (2001, filled triangle), Gorjian, Wright, & Chary (2000, open squares), and Wright & Reese (2000, filled square). The lines which indicate models are all labeled and are from Malkan & Stecker (1998, MS98), Dwek et al. (1998, D98), and Fall, Charlot, & Pei (1996, FCP96) as described in Section 6. The shaded region shows the D98 model rescaled to match the range allowed by our EBL detections.
2 The total energy per unit increment of wavelength is given by I = I d = I d ln. By plotting energy as I = I against log, the total energy contained in the spectrum as a function of wavelength is proportional to the area under the curve. We give I in standard kms units of nW m-2sr-1, equivalent to 10-6 ergs s-1cm-2sr-1. Back.