4.3. Discussion

Evolution in the luminosity density of the form (1 + z)⁽⁾ at 0 < z < 1 and slower growth or stabilization at z > 1, such as suggested by redshift surveys at 0 < z < 4, is consistent with the detected EBL for values of () consistent with CFRS. The strongest constraints we can place on the EBL span a factor of 5 in flux. As such, stronger evolution between 0 < z < 1 than reported by CFRS or continuing evolution at z > 1 cannot be tightly constrained. We note that recent results from Wright (2001), which constrain the 1.25µm EBL flux to be 2.1(± 1.1) × 10^-9 cgs, are in good agreement with our results, but do not improve the constraints on the high redshift optical luminosity density over those discussed above.

In contrast with our results, previous authors have claimed good agreement between the flux in the raw number counts from the HDF and integrated flux in the measured CFRS luminosity density to z = 1 under the assumption that the luminosity density drops rapidly at z > 1 (Madau, Pozzetti, & Dickinson 1998, Pozzetti et al. 1998). In that work, the errors in faint galaxy photometry which cause ~ 50% underestimates of the total light from V > 23 galaxies (discussed in Section 4.1) are compensated by the assumption that the luminosity density drops rapidly beyond z = 1. That assumption was based on measurements of the flux from Lyman-limit systems in the HDF field by Madau et al (1996), which are substantially lower than measurements by Steidel et al. (1999) due to underestimates of the volume corrections and to the small-area sampling. We find that the detected EBL is not consistent with luminosity evolution comparable to the CFRS measured values at 0 < z < 1 if the luminosity density drops rapidly at z > 1.