next up previous contents
Next: Resampling and checking the Up: Wavelength Calibration Previous: Estimating the angle of

Identification loop

The identification criterion combines estimate of the wavelength of a line and an estimate of the error as well as a list laboratory wavelengths to determine and guarantee the identification of a given line. One could involve the additional knowledge of line strengths. But in practice this information of limited use, in reason of variations of relative line intensities caused by impurities, lamp ageing, pressure variations, a.s.o. Accordingly, the identification criterion is:

$ \lambda_c \equiv \lambda_{cat}$ if $\exists$! $\lambda_{cat}$ / $\parallel
\lambda_c - \lambda_{cat} \parallel <$ $\delta\lambda$

with $\lambda_c$, computed estimated wavelength, $\lambda_{cat}$the catalog wavelength and $\delta\lambda$, a majorant of the error of the computed wavelength, taken as the distance of the closest neighbor in the line catalog or in the arc spectrum divided by a coefficient $\alpha$ which value does not exceed 1 (this parameter is controlled by the session keyword WLCLOOP(2)).



http://www.eso.org/midas/midas-support.html
1999-06-15