MEAN-FLUX-REGULATED PRINCIPAL COMPONENT ANALYSIS CONTINUUM FITTING OF SLOAN DIGITAL SKY SURVEY Lyα FOREST SPECTRA

Continuum fitting is an important aspect of Ly alpha forest science, since errors in the derived optical depths scale with the fractional continuum error. However, traditional methods of estimating continua in noisy and moderate-resolution spectra (e.g., Sloan Digital Sky Survey, SDSS; S/N less than or similar to 10 pixel(-1) and R 2000), such as power-law extrapolation or dividing by the mean spectrum, achieve no better than similar to 15% rms accuracy. To improve on this, we introduce mean-flux-regulated principal component analysis (MF-PCA) continuum fitting. In this technique, PCA fitting is carried out redward of the quasar Ly alpha line in order to provide a prediction for the shape of the Ly alpha forest continuum. The slope and amplitude of this continuum prediction is then corrected using external constraints for the Ly alpha forest mean flux. This requires prior knowledge of the mean flux, < F >, but significantly improves the accuracy of the flux transmission, F equivalent to exp(-tau), estimated from each pixel. From tests on mock spectra, we find that MF-PCA reduces the errors to 8% rms in S/N similar to 2 spectra, and <5% rms in spectra with S/N greater than or similar to 5. The residual Fourier power in the continuum is decreased by a factor of a few in comparison with dividing by the mean continuum, enabling Ly alpha flux power spectrum measurements to be extended to similar to 2x larger scales. Using this new technique, we make available continuum fits for 12,069 z > 2.3 Ly alpha forest spectra from SDSS Data Release 7 for use by the community. This technique is also applicable to future releases of the ongoing Baryon Oscillations Spectroscopic Survey, which obtains spectra for similar to 150,000 Ly alpha forest spectra at low signal-to-noise (S/N similar to 2).