New Ar-Ne lamp: blending of line 6678.276 Ne I

[Paolo Berardi]

I'd like to report a blending for one of reference Ne I lines near H-alpha. It is normally used in the standard spectra reduction with 1200 l/mm grating. In ISIS is taken into account by "Predefined mode 1200 groves/mm (6 lines)".

At this resolution I see a blend of 6678.276 Ne I with 6677.281 Ar I (lambda from NIST):



A confirmation of the problem comes from RMS value, larger using predefined mode.

Predefined mode 1200 groves/mm (6 lines):

Spectral calibration
Coefficient a4 : 0.000000E00
Coefficient a3 : -4.244473E-09
Coefficient a2 : 4.601497E-06
Coefficient a1 : 0.35143
Coefficient a0 : 6335.730
--------------------------------------------------------------
RMS : 0.173095

File mode excluding the line Ne I 6678.276 (and adding the Ar I 6752.834):

Wavelength file : C:\Astronomia\Spettroscopia....\L1200_ha.lst
Coefficient a4 : 0.000000E00
Coefficient a3 : -1.811086E-10
Coefficient a2 : -3.135225E-06
Coefficient a1 : 0.35546
Coefficient a0 : 6335.198
--------------------------------------------------------------
RMS : 0.006629

I don't know if this blend happen in every Ar-Ne lamps (it should!) and probably it is not a problem in low-res spectra calibration, anyway I wanted to report it.

Paolo

[Robin Leadbeater]

Good spot Paolo !

I do not think the Ar 6752.834 line will fit in the field of the ICX285AL CCD but I can choose the Ne 6717.043 line instead when I upgrade my LHIRES. This would also work with either Ne or Ne Ar lamps.

Thanks!
Robin

EDIT: I see that Ne 6717.043 is one of the lines already used in the predefined 6 lines mode so it might be better for me to leave the blended line out of the fit and use just 5 lines

[Robin Leadbeater]

I see this blended line is also used in the 4 line mode for the LHIRES 2400 where it will cause a larger error. Unless the ratio of Ne/Ar is constant and a new line wavelength can be calculated, the ISIS LHIRES 2400 4 line calibration mode for H alpha probably should not be used.

I suggest posting a link to this thread on the other spectroscopy forums to avoid potential calibration errors using the Ne Ar lamp when calibrating H alpha spectra at high resolution

Robin