Abstract

We present a method to make predictions with sets of correlated data values, in this case QSO flux spectra. We predict the continuum in the Lyα forest of a QSO, from 1020 to 1216 A, using the spectrum of that QSO from 1216 to 1600 A. We find correlations between the unabsorbed flux in these two wavelength regions in the Hubble Space Telescope (HST) spectra of 50 QSOs. We use principal component analysis to summarize the variety of these spectra, relate the weights of the principal components for 1020-1600 A to the weights for 1216-1600 A, and apply this relation to make predictions. We test the method on the HST spectra and find an average absolute flux error of 9%, with a range of 3%-30%, where individual predictions are systematically too low or too high. We mention several ways in which the predictions might be improved.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call