Hi Gary and Ryan,

Thanks for your help and insight. I respond below to Gary's statements, with a
remaining question at the end.

--- In cloudy_simulations@yahoogroups.com, "gary_ferland" <gjferland@...> wrote:
>
> Hi Kirk,
>
> here are a few things that I think are happening.
>
> in its default state the code does FeII with the lowest 16 levels as in the
real atom, then the Wills, Netzer & Wills collapsed atom, finally Lya -> FeII
pumping from something Fred Hamann put together long ago, but which was never
published.
>
Ok, yes, that's about what I thought.

> When the size of the atom is increased the logic does not check to what extent
it was increased - it turns off the approximations and lets the FeII atom stand
on its own. The intermediate sized atom does not have enough high transitions
to pick up Lya at all,

Ok, that jives with what I had written in my last post.

>but even the full 372 levels do not get all of Lya - the levels which are most
directly pumped are higher than we have. There are no collision data for the
important very high levels.
>

Yep, a nasty fact of life for now.

> The feedback you see is very physical - Lya is strongly trapped with little
net emission. The 2p level is depopulated by absorption of Lya. That is by
photoionization of 3rd row elements, and of H 2s&p and HeI n=2. These
populations are strongly affected by J-bar in Lya. And J-bar in Lya is affected
by the other species, and by FeII.
>

Ok. I understand (mostly -- see final question, below) the origin of the
behavior that we were seeing, although it was the "cliff" effect for the default
FeII model atom case that alerted me to something possibly going awry.

If I understand correctly, the overall effects of the trapped Lya on the HeI*
and neutral heavies is physical, but that the cliff behavior for the default
FeII model atom case is probably an artifact of the simplicity of that model.

> The fact is that if we had atomic data for the much higher levels the
populations would change some more. In the limit where these other species
remove all Lya you would get the "no ots" answer.
>

This is the remaining piece that still puzzles me -- why aren't the heavy
neutrals affected by Lya photoionization when the Lya pumping of FeII isn't
considered in the 99-level atom case? (note the much larger ion fractions of the
neutrals for this model) Naively, I would have thought that they'd be clobbered
even further, given that Lya wasn't being destroyed by pumping FeII. (?)


> Most BLR results are affected by turning on the large FeII atom. The simple
approximations simply do not recover its rich physics. The same is true for
H_2.
>
> good luck,
> Gary
>
yep, and thanks again.