Quasar Composite Spectra and Spectrophotometric Connection

Quasar Composite Spectrum

The method used to create a quasar composite spectrum with 102,150 BOSS quasars is fully described in . Here, the process is summarized.

The quasar sample used to create the composite is limited:
To quasars with redshift between 2.1 and 3.5,
By using quasars targeted by XDQSO and/or BONUS,
By excluding BALs and DLAs,
And by excluding quasars observed at an airmass greater than 1.2.

These sample selections result in using 102,150 quasar spectra to generate the composite.

A spectrophotometric correction, a function of wavelength and airmass, is applied. This correction is described in detail in Harris et al 2015 and described briefly . An optical depth correction is also applied.

To assemble the composite, the individual spectra were normalized. The spectra were ordered by redshift and normalized 500 at a time to reduce computing time while still generating a useful composite.

Spectrophotometric Correction

The method used to create the spectrophotometric correction used for the quasar composite spectrum (link to paper) is fully described in that paper. Here, the process is summarized.

The spectrophotometric correction is generated by comparing the spectra of 4000 Å standard stars reduced through a calibration using 4000 Å standard stars against the spectra of those same standard stars reduced through a calibration using 5400 Å standard stars. The mean flux ratio at each wavelength on each plate is recorded. At each wavelength, these flux values for each plate are linearly fit with a constraint that the fit must pass through a ratio of one at an airmass of one. The slope of this fit generates the correction at each pixel.

The correction is applied as follows:

Fc(λ) = F(λ)S(λ)X + I(λ)

where F_c is the corrected flux, F is the uncorrected flux, S is the slope, X is the airmass, and I is the intercept.