A second pass in the optics of the PACS spectrometer can cause a ghost image on some spaxels. Figure 4.23 shows the spaxels where a second pass ghost might appear, and the location of the corresponding spaxels where the originating, real emission is located. The ghost appears shifted in wavelength. If a source in one of the originating spaxels shows a strong spectral line, typically an atomic fine structure line, a weak, broadened line can be seen at an offset wavelength in the corresponding spaxel affected by 2nd pass ghosts. The peak flux of this line is typically ~5% of the line peak of the originating line. The integrated line flux can be up to ~14% of the integrated line flux of the originating line. An example is shown in Figure 4.24. The wavelength offset between the originating line and the ghost line depends on the spectral order of the band, and varies with wavelength. A few examples for the strongest fine structure lines in the PACS wavelength range are given in Table 4.2. A more comprehensive list of frequently observed lines is provided in the PACS Spectrometer Calibration Document Appendix B.

Before interpreting broad spectral lines in spaxels potentially affected by the 2nd pass ghosts, observers should use these tables and, if available, the spectra observed in the corresponding ghost source spaxel, for the presence of a strong line at the originating wavelength. Point source observations, well centered on the central spaxel, are not affected. The central spaxel of the IFU is not contaminated by 2nd pass ghosts.