Tentative detection of the cosmic infrared background at 2.2 and 3.5 microns using ground-based and space-based observations

The cosmic infrared background (CIRB) is the sum total of the redshifted and reprocessed short-wavelength radiation from the era of galaxy formation, and hence contains vital information about the history of galactic evolution. One of the main problems associated with estimating an isotropic CIRB in the near-infrared (1-5 mu m) is the unknown contribution from stars within our own Galaxy. The optimal observational window to search for a background in the near-IR is at 3.5 mu m since that is the wavelength region where the other main foreground, the zodiacal dust emission, is the least. It is not possible to map out the entire 3.5 mu m sky at a resolution that will accurately estimate the flux from stars. However, since the CIRB is presumably isotropic, it can potentially be detected by selecting a smaller field and imaging it at good resolution to estimate the stellar intensity. We selected a 2 degrees x 2 degrees dark spot near the north Galactic pole which had the least intensity at 3.5 mu m after a zodiacal light model was subtracted from the all-sky maps generated by the Diffuse Infrared Background Experiment (DIRBE). Still, the large area of the held made it very difficult to mosaic at 3.5 mu m using the available arrays. Thus, the held was mosaicked at 2.2 mu m, then the bright stars were selected and reimaged at 2.2 and 3.5 mu m. The resulting total intensity of the bright stars was combined with a model for the contribution from dimmer stars and subtracted from the zodi-subtracted DIRBE map. The contribution from the interstellar medium was also subtracted, leaving a residual intensity at 2.2 mu m of 16.4 +/- 4.4 kJy sr(-1) or 22.4 +/- 6.0 nW m(-2) sr(-1), and at 3.5 mu m of 12.8 +/- 3.8 kJy sr(-1) or 11.0 +/- 3.3 nW m(-2) sr(-1). The nature of our analysis suggests that this excess emission is probably a detection of the cosmic background in the near-infrared.