Initial baryon temperature of small_cosmo_volume example seems wrong

[apontzen]

In small_cosmo_volume.yml, the initial temperature of the baryons at z~=50 is set to ~7000K. An accompanying comment says (1 + z_ini)^2 * 2.72K.

This temperature seems far too high, and the comment makes it sound like it's a mistake rather than intentional. Baryons thermally decouple from the CMB photons at z~150, and only after that time would their temperature redshift as (1+z)^2. Thus, a more appropriate temperature would seem to be T_cmb (1+z)^2 / (1+z_dec) which comes to ~45K, below the temperature floor for the simulation.

Other examples give more reasonable initialisation temperatures, e.g. eagle_50.yml initialises to T=268.7K at z=127, which seems sensible. Given that it is quoted to 4 significant figures, perhaps it's based on a more detailed calculation than the above hand-waving (which would produce a slightly higher estimate).

I have never played much with changing initialisation temperatures. I would assume the effect is fairly modest, but such an extreme discrepancy in the provided small_cosmo_volume.yml example seems like it could generate unphysical thermal smoothing effects at early times. Is there a specific reason for it?