I have two questions about analysis of 1D-spectra in IRAF
Danilenko Andrey wrote on Sep 10, 2006
Hello everybody!
I have two questions about analysis of 1D-spectra in IRAF:
- first one is how to examine quality of flux calibtration or how to estimate error (or rms; not formal but real rms) of flux calibration?
- second one is how to verify identified emission lines (I run IDENTIFY task to identify this lines)?
Help me to solve this two problems.
Thanks for your attention,
Andrey.
I have two questions about analysis of 1D-spectra in IRAF:
- first one is how to examine quality of flux calibtration or how to estimate error (or rms; not formal but real rms) of flux calibration?
- second one is how to verify identified emission lines (I run IDENTIFY task to identify this lines)?
Help me to solve this two problems.
Thanks for your attention,
Andrey.
Francisco Valdes wrote on Sep 10, 2006
Hello Andrey,
The best way to empirically estimate the uncertainty in a spectrum, flux calibrated or not, is the rms in a continuum region. The RMS measurement should be about a fitted continuum unless the continuum is pretty flat. Note that this will give the uncertainty of a pixel value. It will not give you the uncertainty in the actual flux calibration, which is maybe why you are asking. In this case you would have to look at the uncertainties in the response points from STANDARD. So this is not an easy question and I don't have an complete answer.
I am not sure what you mean by "verify dentified emission lines". If you mean whether you have assigned the correct wavelength to an emission line then there are two things to look at. One is a literature or finding chart. For some common arc lamps I can point you to: http://www.noao.edu/kpno/specatlas/. The other thing that is generally done is to assume you have most of the lines correctly identified. Then when you fit the dispersion function misidentified lines should have large residuals. You can then delete or iteratively reject the lines to find a good dispersion function. The you can add lines based on this dispersion function to get you closer to the correct line.
I hope this helps.
Yours,
Frank Valdes
The best way to empirically estimate the uncertainty in a spectrum, flux calibrated or not, is the rms in a continuum region. The RMS measurement should be about a fitted continuum unless the continuum is pretty flat. Note that this will give the uncertainty of a pixel value. It will not give you the uncertainty in the actual flux calibration, which is maybe why you are asking. In this case you would have to look at the uncertainties in the response points from STANDARD. So this is not an easy question and I don't have an complete answer.
I am not sure what you mean by "verify dentified emission lines". If you mean whether you have assigned the correct wavelength to an emission line then there are two things to look at. One is a literature or finding chart. For some common arc lamps I can point you to: http://www.noao.edu/kpno/specatlas/. The other thing that is generally done is to assume you have most of the lines correctly identified. Then when you fit the dispersion function misidentified lines should have large residuals. You can then delete or iteratively reject the lines to find a good dispersion function. The you can add lines based on this dispersion function to get you closer to the correct line.
I hope this helps.
Yours,
Frank Valdes
Last post on Sep 10, 2006