Filling Gaps in Indo-US Stellar Spectral Library using Principal Component Analysis

Abstract

The Indo-US coud´e feed stellar spectral library (CFLIB) published recently by Valdes et al. (2004) contains spectra of 1273 stars in the spectral region 3460 to 9464 ˚A at a resolution of 1 ˚A. About 500 stars in this database have gaps ranging from a few ˚A to several tens of ˚A in this wavelength range. We use a variation of Principal Component Analysis (PCA) technique to fill gaps of up to 5˚A in a subset of spectra from the CFLIB. We hope to exploit the full potential of the scheme and attempt to fill larger gaps in stellar spectra in a subsequent study.