splot’s measurements

Sergei O. Naoumov wrote on Oct 15, 1998

Hi!
I looked at the splot's help file and it says that when hitting 'e'
keystroke that is supposed to just add up the pixels to measure the
equivalent widths, it uses:

	 eq. width = sum (1 - I(i)/C(i))
	 
But the equivalent width in this formula depends on how many points
one wants to add up! The question is what exactly is splot doing
then?

Thanks,
		Sergei

---

Frank Valdes wrote on Oct 20, 1998

> In article <705b1k$878@tucana.tuc.noao.edu>, "Sergei O. Naoumov" <serge@avarice.astro.unc.edu> writes:
> From: "Sergei O. Naoumov" <serge@avarice.astro.unc.edu>
> Subject: splot's measurements
> 
> 
> Hi!
> I looked at the splot's help file and it says that when hitting 'e'
> keystroke that is supposed to just add up the pixels to measure the
> equivalent widths, it uses:
> 
> 	 eq. width = sum (1 - I(i)/C(i))
> 	 
> But the equivalent width in this formula depends on how many points
> one wants to add up! The question is what exactly is splot doing
> then?
> 
> Thanks,
> 		Sergei


Hello Sergei,

The formula is correct appart from a multiplication by the wavelength
per pixel to convert the integration over pixels to an integration over
wavelength and to give the equivalent width the units of wavelength.

Yes the answer depends on the number of pixels used.  But in an ideal
isolated line profile the line intensities I(i) converge on the continuum
C(i) as you move away from the line center.  Thus the quantity being
summed becomes zero and so the equivalent width converges to a value
which is independent of the region being summed.  In practice in SPLOT
you mark the region to be summed to be points where the profile reaches
the continuum.

Cheers,
Frank Valdes

---

Last post on Oct 20, 1998