New limits on the density of the extragalactic background light in the optical to the far infrared from the spectra of all known TeV blazars

Aims. The extragalactic background light (EBL) in the ultraviolet to far-infrared wavelength region carries important information about galaxy and star formation history. Direct measurements are difficult, especially in the mid infrared region. We derive limits on the EBL density from the energy spectra of distant sources of very high energetic.gamma-rays ( VHE.gamma-rays). Methods. The VHE.gamma-rays are attenuated by the photons of the EBL via pair production, which leaves an imprint on the measured spectra from distant sources. So far, there are 14 detected extragalactic sources of VHE.gamma-rays, 13 of which are TeV blazars. With physical assumptions about the intrinsic spectra of these sources, limits on the EBL can be derived. In this paper we present a new method of deriving constraints on the EBL. Here, we use only very basic assumptions about TeV blazar physics and no pre-defined EBL model, but instead a large number of generic shapes constructed from a grid in EBL density vs. wavelength. In our study we utilize spectral data from all known TeV blazars, making this the most complete study so far. Results. We derive limits on the EBL for three individual TeV blazar spectra ( Mkn 501, H1426+ 428, 1ES 1101- 232) and for all spectra combined. Combining the results from individual spectra leads to significantly stronger constraints over a wide wavelength range from the optical (similar to 1 mu m) to the far-infrared (similar to 80 mu m). The limits are only a factor of 2 to 3 above the absolute lower limits derived from source counts. In the mid-infrared our limits are the strongest constraints derived from TeV blazar spectra so far over an extended wavelength range. A high density of the EBL around 1 mu m, reported by direct detection experiments, can be excluded. Conclusions. Our results can be interpreted in two ways. ( i) The EBL is almost resolved by source counts, leaving only a little room for additional components, such as the first stars, or ( ii) the assumptions about the underlying physics are not valid, which would require substantial changes in the standard emission models of TeV blazars.