Heating rates in Cloudy _shielded tables are too high for low density and low temperature gas

[AnnaGenina]

Low-density and low-temperature gas is overheated with the _shielded tables, which is also very odd because shielding is not expected to have any effect in this regime. There is no issue with the Cloudy tables that do not include shielding.

As an example of what kind of impact this can have, I am showing here the evolution of gas temperature at mean cosmic density as a function of redshift for a cosmological simulation that only includes gas cooling (no star formation, no feedback). It is clear that the shielded tables overheat this low-density gas.

Thanks to Volker Springel, Yves Revaz, and Matthew Smith for discovering and investigating this problem.

[mcsmithastro]

I have been looking at the values in the cooling tables directly and comparing to the unshielded tables. I find that there are certainly strange things in the primordial rates and probably in the metal rates too. FG2011 is much worse than HM2012, but they are both concerning. I have made a series of plots which you should be able to see here: https://datashare.mpcdf.mpg.de/s/fVuWbDrsCSD3dLx

For the FG2011 UVB, the fg2011_heat (fg2011_cool) directory contains plots of the ratio of the primordial heating (cooling) rates in the shielded vs unshielded files. Each plot corresponds to a different redshift and shows the ratio as a function of temperature for different densities. fg2011_abs shows absolute value of (heating - cooling) with thick lines indicating the shielded and thin lines showing the unshielded tables, solid lines indicating net cooling and dashed lines indicating net heating. fg2011_met_* show the same set of plots for the metal tables (not including the primordial contribution). hm2012* show the same plots for the HM2012 UVB.

The takeaways for primordial FG2011: Excess heating in IGM density gas relative to the unshielded tables by up to a factor of 1e7 at most redshifts at all temperatures. There are frequently artefacts where there is abruptly an ehancement of heating for a specific density, temperature, redshift combination which then disappears for adjacent regions of the parameter space. Several elements of the arrays set to zero, despite being surrounded by non-zero values in all directions.

Takeaways for primordial HM2012: Not as bad as FG2011, but excess net heating relative to the unshielded tables for gas below 2e4 K for IGM density gas at most redshifts by up to a factor of 1000. This stems from an abrupt enhancement in the heating rates at 2e4 for gas denser than 100 H/cc, but only leads to a net heating enhancement at IGM densities.

The metal rates are much harder to interpret as they are much more complicated as a function of (rho,z,T). It is less clear cut whether there is something wrong but there are a few features I find concerning, both in IGM and near-ISM density gas. Also, given that the primordial rates look so odd I am not sure if we can trust the metal rates that presumably originate from the same cloudy runs. Again, FM2011 shows very strong net heating relative to the unshielded tables at all temperatures and most redshifts for IGM density gas. Frequently there is significant net heating in gas around 0.1 H/cc and 1e2 - 1e4 K where the unshielded tables contain net cooling. Again, there are many zero elements. It is a similar picture for HM2012, but less severe.

[brittonsmith]

@gregbryan, do you have any thoughts about this?