question regarding spectra

[Jonathan Slavin]

Hi all,

I'm wondering about the interpretation of the intensity returned by the spectrum function. Should the value be interpreted as applying to each wavelength point or is it a binned value over the wavelength bin? This is relevant for calculations, e.g. of the integration of the spectrum times effective area for a detector over some wavelength range. If the spectrum applies to the point, then one would want to use something like Simpson's method, whereas if it's binned then one would simply sum *dwvl where dwvl is the bin width.

Thanks,

Jon

[Ken Dere]

Hi Jon,

for spectra, a filtering function is applied, such as a gaussian.  the default is a gaussian (guassianR) with a resolving power R of 1000. so the std is wvl/R.  In these cases you would want to integrate over the calculated line profile where dwvl is obtained from the input wavelengths.  So, in a sense, they are binned.

The spectral line intensities are also available as the attribute Intensity (a dict) from the spectral calculations.

One could also use the bunch class that does not apply a filter but also does not calculate the continuum

hope this helps.

Ken