10.2. Discussion: Minimum EBL23
Using the ensemble photometry method described above, we measure the flux from detectable sources relative to the mean sky level in the F300W, F555W, and F814W images of our own data to be 3.19 × 10-9, 6.02 × 10-10, and 5.16 × 10-10, respectively. (5) For the F300W and F814W images, we use the masks derived from the F555W images in order to guarantee that the same sources are contributing to the minimum EBL23 at all wavelengths. The 1800 galaxies detected in the HDF with V606 magnitudes in the range 27.5-30 A B mag (see Figure 8) are clearly not included in our estimate of the minimum EBL23 derived by ensemble photometry.
The contribution from galaxies in the HDF catalog with V555 > 27.5 A B mag is 0.57 × 10-10ergs s-1 cm-2 sr-1 Å-1 at V555 as measured by standard photometric methods. In keeping with the discussion in the previous section, we estimate that only 35% of the light is recovered from galaxies with 27.7 < V606 < 30 A B mag in the HDF, so that the true flux from these sources is roughly 1.8 × 10-10ergs s-1 cm-2 sr-1 Å-1. Adding this to the flux from detected sources (23 < V555 < 27.5) in our EBL field, we find the total flux from detected sources (23 < V555 < 30) to 4riso apertures to be 7.8 × 10-10ergs s-1 cm-2 sr-1 Å-1. Adding to this the estimated extragalactic contribution to the background sky level beyond 4riso (Figure B2), we identify the minimum flux from detected galaxies with V555 > 23 A B mag to be 8.9 × 10-10ergs s-1 cm-2 sr-1 Å-1. We emphasize that this estimate of the minimum EBL23 is indeed a minimum estimate from which sources will be excluded due to the surface brightness biases and detection limits of our own images as well as the HDF (see Figures 19 and 20). Incompleteness due to surface brightness detection limits is discussed further in Paper III.
Figure 20. The same quantities as plotted in Figure 19 are shown for the galaxies in the HDF F606W catalog (Williams et al. 1996). The detection limit of the HDF catalog for V606 is quoted as 28.2 AB mag, which corresponds to a 10 detection in the drizzled images (~ 2sky in 4 pixels in the original images). At V606 = 28.2AB mag, a typical galaxy has (µiso - µcore) 1 mag arcsec-2. See Appendix B.2 for discussion.
Following the same method for F300W and F814W data leads to the minimum EBL23 values summarized in Table 10. The total combined statistical and systematic error for this minimum EBL23 measurement is roughly ± 1 × 10-11 ergs s-1 cm-2 sr-1 Å-1, × 100 smaller than the error for the EBL23 detections because the conversion to physical units can occur after foregrounds are subtracted. Comparing the total flux from detected galaxies measured using ensemble photometry versus standard methods, the photometry errors affecting standard methods are clearly worse at UV wavelengths where signal-to-noise ratio is lower, calibration is less accurate, and the instrumental PSF is broader than in the other filters.
5 For comparison, the flux measured by standard photometry for sources with AB magnitudes between 23 and 27.5 in the HDF are 6.7, 5.1, 3.8, and 2.5 × 10-10ergs s-1 cm-2 sr-1 Å-1 in the F300W, F450W, F606W, and F814W, respectively. Back.