As part of WAVES redshifts will be automatically assigned to the 4MOST reduced spectra on a weekly basis, and the WAVES input catalogue will be updated to remove sources for which a secure redshift has been obtained. This process will be achieved using an adapted version of the auto-z redshift fitting code developed for the Galaxy And Mass Assembly survey by Ivan Baldry. Full details of the autoz code can be found in Galaxy And Mass Assembly (GAMA): AUTOZ spectral redshift measurements, confidence and errors MNRAS, 2014, 441, 2440 and an IDL implementation of the code can be found on Ivan Baldry’s home page .
Briefly, auto-z continuum-subtracts the target spectra and cross-correlates with a sample of continuum-subtracted galaxy templates to identify both absorption-line and emission-line features. In this process, deviations in the high-pass filtered spectra are clipped to exclude uncorrected artifacts and to reduce the significance given to single-line matches. auto-z then outputs best-fit redshifts and confidences, and assigns a likelihood probability to each redshift.
The code have been adapted to run in the [R] programming language and will be further adapted for the specifics of the 4MOST data outputs. A version is currently running on the spectral simulator tool SpecGen, allowing the user to simulate an observed spectrum and ascertain whether are redshift would be obtained when using auto-z.
For further information contact: luke.j.davies@uwa.edu.au