Lyman-alpha forest: why not Lyman-beta?

[astrolollo]

Hello everyone!
We observe the so-called Lyman-alpha forest in the spectrum of distant quasars and it is said that these multiple absorption lines are due to the presence of intergalactic HI clouds that absorb light at the wavelength of 1216 A , the Lyman-alpha transition. My question is the following. Why are these absorption lines referred only to Lyman-alpha processes and not, for example, to Ly-beta transitions? Thanks to everyone who will answer.
Lorenzo

 

[BvU]

Perhaps because of the intensity ratio between Ly$\beta$ and Ly$\alpha$ ?
(about a factor 100 less for the solar spectrum -- don't know about distant quasars)

 

[astrolollo]

I don't think this is the case.. The high-z quasar has a continuum emission. A cloud at lower z will not interact with the radiation coming from the Ly-a or Ly-beta emission of the quasar, but with photons that were emitted at a higher energy. Since the cloud should be quite cold (~10 K ?) all the hydrogen is in the fundamental state. As far as I know, if a gas of atoms in the fundamental state is irradiated with light, there shouldn't be any privileged transition . That is, Ly-a is as likely as Ly-beta or Ly-gamma. Am I wrong? Thanks.

 

[Bandersnatch]

Am I wrong?

See the explanation here:
https://astrobites.org/2013/07/14/astrophysical-classics-neutral-hydrogen-in-the-universe-part-1/
especially the animations.

 

[BvU]

Strong argument. My mistake (switching emission and absorption) is embarassing ...
We need an expert. @Andy Resnick, perhaps ?
PS what about the last figure here ?

[edit] ahhh, BS to the rescue ! To me it looks like a mixed $\alpha,\beta$ forest ?
the picture in the wiki lemma also shows absorption 'peaks' (at the red arrows) -- for much longer wavelengths -- with the same $\lambda_\alpha/\lambda_\beta$ ratio

(Wiki says: Picture ESO)

 

[Ken G]

Yes, there should be Lyman beta as well, but the cross section for absorption is much lower, perhaps a factor of about 5 or a little more if I recall correctly. That creates a proportionally less deep absorption line, if the Lyman alpha is not saturated (i.e., not completely dark at line center). It has to do with the quantum mechanical coupling between the ground state and the second and third levels, the former is quite a bit stronger than the latter so you get deeper absorption in Lyman alpha. Still, I imagine that any time you see a Lyman alpha line in the spectrum, and you can infer the hydrogen column producing it, you immediately include the corresponding weaker absorption feature at the wavelength of Lyman beta. If it's not seen in the spectrum, you have a problem! Also, note that since the wavelength ratio is (1-1/4)/(1-1/9) = 27/32, then for any observed quasar at redshift z, you have a "clean" purely Lyman alpha region from that z down to a z of 27/32*(z+1) - 1, which is 27z/32 - 5/32. It seems you could model the Lyman alpha absorption in that region, calculate the corresponding Lyman beta absorption and remove it from the data at lower wavelengths, then what's left at the lower wavelengths is again all Lyman alpha, which you model and remove the Lyman beta at the next lowest series of wavelengths, etc.