The proximity effect on the Lyman α forest due to a foreground QSO

The influence of a foreground QSO on the lyman a forest of another QSO withhigher redshift has been investigated by analyzing the spectra of three suchobjects at redshifts $z=2 - 2.7$. This influence is not contaminated by anyprojection effects, as opposed to the inverse effect along the line of sight,where incomplete coverage of the QSO continuum emitting region by the lyman aclouds could contribute to the relative lack of lines. Our results areconsistent with the existence of a proximity effect due to the foreground QSO,but due to its weakness we can only reject the absence of such effect at 1sigma level. By modelling this proximity effect assuming that lyman a cloudsare low-density highly ionized objects we find that the best value for the UVphotoinizing intensity at those redshifts is 10^{-20.5}(...) at the Lyman limitwhich is consistent with previous estimates of the background based on theinverse effect. We also find an absolute lower limit (at 95% confidence) to theUV intensity at the level of 10^{-21.8}(...) which means the rejection of anumber of models for the UV background.Comment: 10 pages + tar file with 6 figures and 4 pages of tables, uses MNRAS style file