Files accompanying Brodzeller et al. (2023). Included in this repository are the quasar templates, SDSS redshift catalog used for training the templates, and an updated HIZ template that accounts for lya mean transmission. The latter model is discussed in Bault et al. (2024). All template files are compatable with redrock. LOZ/HIZv1.0 produced the standard quasar classifications for DESI year one data release. HIZv1.1 refined the z>1.6 DESI year one redshifts for Lya forest cosmology studies, redshift catalog to be released TBD.
Contact abrodze@gmail.com for questions about this repository.
rrtemplate-qso-LOZ.fits - low redshift (0.05<z<1.6) quasar templates presented in Brodzeller+23
rrtemplate-qso-HIZv1.0.fits - high redshift (1.4<z<7) quasar templates presented in Brodzeller+23
rrtemplate-qso-HIZv1.1.fits - Upgraded high redshift (1.4<z<1.7) quasar templates, trained on identical sample as above HIZ but individual spectra were corrected for Lya effective optical depth via Kamble et al. (2020); redshift performance described in Bault et al. (2024)
trainingsample.fits - catalog of SDSS spectra used to train the above templates; data model described below
clusters directory - the low-redshift and high-redshift subdirectories contain cluster information corresponding to Table 2 of Brodzeller et al. (2023). Cluster information for the 0<z<0.35 redshift bin is not included - please email me for the results of this bin. There are four files per redshift bin:
- clustersLSNR_z{zmin}-{zmax}.fits.gz : SDSS plate-mjd-fiber-Z_PCA of quasar spectra within each cluster of the low signal-to-noise subset, cluster # (extension name) corresponds to its index in FLUX/IVAR extension of its corresponding composite_spectra*.fits
- composite_spectraLSNR_z{zmin}-{zmax}.fits.gz : composite spectra constructed from the clusters in corresponding above file, re-sampled to shared loglambda wavelength (extension 'LOGLAMBDA') array
- clustersHSNR_z{zmin}-{zmax}.fits.gz : SDSS plate-mjd-fiber-Z_PCA of quasar spectra within each cluster of the high signal-to-noise subset, cluster # (extension name) corresponds to its index in FLUX/IVAR extension of composite_spectra.fits
- composite_spectraHSNR_z{zmin}-{zmax}.fits.gz : composite spectra constructed from the clusters in corresponding above file, re-sampled to shared loglambda wavelength (extension 'LOGLAMBDA') array
To plot the individual eigenspectra from each file in python:
from astropy.io import fits
import numpy as np
import matplotlib.pyplot as plt
with fits.open('rrtemplate-qso-LOZ.fits') as h:
wave = 10**(h[0].header['CRVAL1'] + np.arange(h[0].header['NAXIS1'])*h[0].header['CDELT1']) #wavelength stored in log lambda format
vectors = h['BASIS_VECTORS'].data #eigenspectra
#redshifts = h['REDSHIFTS'].data #redshift coverage of model for running redrock
for i in range(4):
plt.plot(wave, vectors[i]/np.median(vectors[i]), label=f'vector #{i}')
plt.xlabel(r'rest-frame wavelength [$\AA$]')
plt.ylabel('flux density')
plt.legend()
plt.show()
| Column name | # | Description |
|---|---|---|
| plate | 0 | SDSS spectroscopic plate number |
| mjd | 1 | Modified Julian day of the spectroscopic observation |
| fiberid | 2 | Spectroscopic fiber number |
| SNR | 3 | mean signal-to-noise ratio per pixel of spectrum calculated using only lambda_RF(Z_PCA) > 1216 AA |
| Z_PCA | 4 | PCA redshift derived by the redvsblue algorithm, provided in the SDSS DR16 quasar catalog |
| Z_FIN | 5 | Redshift used to train LOZ/HIZ quasar templates; derivation described in Section 3.2 of Brodzeller et al. (2023) |
| Z_REFINE | 6 | Redshift determined when Lya forest is masked. If not -1, this redshift is used instead of Z_FIN |