How to validate observations

Forums Spectroscopy How to validate observations How to validate observations

David Boyd

Hello Keith,

Welcome to the exciting (and challenging) world of spectroscopy! You are clearly well-equipped. There are some useful resources on the BAA Equipment and Techniques section webpages at Francois’ book is also a good resource and a good guide for a beginner to get up to speed.

There is an active online discussion group on spectroscopy mainly based in the US but including some of us in the UK which is coordinated by Woody Sims in Phoenix. Several of us now have LISA spectrographs. I will ask Woody to make contact with you. 

It looks from the spectra you have posted that you have some problems with wavelength calibration as the Balmer lines in your spectra do not always lie at the correct wavelengths. I have been using Christian Buil’s ISIS software to process my spectra for a long time but have just started looking at Demetra. It has a good facility for checking how well it is locating the lines in your calibration spectrum so it is worth spending some time working with that. You should be aiming for an rms value of around 0.1 with a LISA. Make sure your LISA is well focused and that you are not saturating the lines in the calibration image. 

So try to master wavelength calibration first then move on to response correction. This involves taking the spectrum of a reference star whose spectrum you can find in a library and using that to generate a smooth response profile which will enable you to correct the spectrum of another star for atmospheric and instrumental losses. The reference star should be close in airmass (=altitude) to your intended target star so the atmospheric correction will be the same. You can find spectra of possible reference stars in the MILES library (these are actual measured spectra) or in the Pickles library (these are synthetic spectra of the main spectral types). There is a useful spreadsheet which will give you a list of possible MILES reference stars close to a specific location in the sky at a specific time here

Try to pick stars which are main sequence stars (spectral class V) and have spectral types B or A as these have spectra with a smooth continuum which makes it easier to generate a smooth response profile in Demetra.

There is also a lot of information available on the ARAS spectroscopy forum at and that is also a good place to ask questions as many active spectroscopists use it.

That’s probably enough information to keep you busy for now. It’s good to have you in the amateur spectroscopy community. There is much useful work to be done.

Best wishes,