[newchem-cpp] new parameters

[brittonsmith]

I'd like to open this as a space for discussion of the various new parameters. I think we are in agreement that there will need to be some reorganization here. I will list them below and follow up with some comments.

We have two new dust models coming and our current treatment of dust. For the sake of discussion I will give them names (also open for debate):

• the "passive" dust model: a dust density is either supplied by the user (use_dust_density_field=1) or assumed to scale with the metallicity. This dust can be used for H2 formation, photo-electric heating, and dust recombination cooling. The field is not evolved in any way.
• the "single-grain" model (i.e., the Harrison model): a single species is evolved with sputtering and accretion and connected to a subset of the new non-eq metal chemistry.
• the "multi-grain" model (i.e., the Gen model): a number, between one and several, specific dust species are evolved and coupled to the full non-eq metal chemistry

New Parameters (currently named)

• metal_chemistry: (0, 1) - turns on non-eq metal chemistry
• grain_growth: (0, 1) - turns on grain growth for
• multi_metals: (0, 1) - 0 to select a single metal yield for dust grains (selected with metal_abundances, 1 to follow 12 supernova progenitors
• metal_abundances: (0-11) - if multi_metals=0, choose one of the options below to follow metals from a single source:
  • 0: metal/dust abundances of local ISM (Pollack et al. 1994)
  • 1-4: Pop III normal core-collapse supernovae (Nozawa et al. 2007) with progenitor masses 13, 20, 25 and 30 Msun
  • 5-8: Pop III faint supernovae (Marassi et al. 2014) with progenitor masses 13, 50 and 80 Msun
  • 9-10: Pop III pair-instability supernovae (Nozawa et al. 2007) with progenitor masses 170 and 200 Msun
  • 11: simple dust model (only include silicate and graphite; Yajima et al. 2017)
• dust_species: (0-3) - enable multlple dust species
  • 0: single species
  • 1: enstatite + amorphous carbon (also follow Mg metal density)
  • 2: + metallic silicon + metallic iron + forsterite + magnetite + silica + magnesia + troilite + alumina (also follow Al, S, Fe metal densities)
  • 3: + water ice + volatile organics + refractory organics
• use_multiple_dust_temperatures: (0, 1) - enable calculating and storing multiple dust temperature (i.e., for each species)
• dust_sublimation: (0, 1) - enable dust_sublimation
• radiative_transfer_HDI_dissociation: (0, 1) - enable photodissociation of HD molecules (adds field: RT_HDI_dissociation_rate)
• radiative_transfer_metal_ionization: (0, 1) - enable photoionization of C and O atoms (adds fields: RT_CI_ionization_rate, RT_OI_ionization_rate)
• radiative_transfer_metal_dissociation: (0, 1) - enable photodissociation of CO, OH and H2O molecules (adds fields: RT_CO_dissociation_rate, RT_OH_dissociation_rate, RT_H2O_dissociation_rate)
• radiative_transfer_use_H2_shielding: (0, 1) - flag to signal H2 self-shielding is being done in hydro code
• hd_reaction_rates: Alternative calculation scheme for k50-k56
• gas_grain_cooling_rate: Alternative calculation scheme for gas_grain
• uniform_grain_isrf_heating_rate: Alternative calculation scheme for gamma_isrf
• use_primordial_continuum_opacity: (0, 1) - include primordial continuum opacity
• h2_cooling_rate: (0-3) - controls H2 cooling rate
  • 0: Lepp & Shull (1983)
  • 1: Galli & Palla (1998)
  • 2: Glover & Abel (2008) DEFAULT
  • 3: Chiaki & Wise (2019)
• hd_cooling_rate: control HD cooling rate
  • 0: Coppola et al (2011) and Wrathmall, Gusdorf, & Flower (2007) DEFAULT
  • 1: Chiaki & Wise (2019)
• tabulated_cooling_minimum_temperature: (float, 10000) - temperature above which we add tabulated metal cooling when metal chemistry enabled

[brittonsmith]

There is a lot to discuss with regard to the above, but my top priority is sorting out the dust and metal chemistry. To that end, I'm proposing something like the following:

• metal_chemistry:
  • 0: off
  • 1: a reduced model to work with the single-grain dust model including, say, CI, CII, OI, OII, CO
  • 2: the full new model
• dust_chemistry:
  • 0: off
  • 1: passive dust model
  • 2: single-grain model
  • 3: multi-grain model

Note, the dust_chemistry aspect breaks current behavior. I'm ok with this. This is somewhat bolder, but there could be an argument for merging metal_cooling into metal_chemistry. Personally, I wonder if we overcomplicate things by distinguishing between chemistry and cooling. For example, we could have:

• metal_chemistry:
  • 0: off
  • 1: tabulated metal cooling (equivalent to metal_cooling=1 now)
  • 2: reduced non-eq model
  • 3: full model

[brittonsmith]

Some updates:

Per PR #599, we now have the following more or less stabilized:

• dust_chemistry:

  • 0: off
  • 1: passive dust model
  • 2: Chiaki multi-grain dust model

• metal_chemistry:

  • 0: off
  • 1: full Chiaki non-equilibrium network

• use_dust_density_field: only works with dust_chemistry = 2

• dust_chemistry = 2 requires dust_species > 0

Parameters removed:

• h2_on_dust (now always on for dust_chemistry > 0 and primordial_chemistry > 1
• dust_sublimation
• use_multiple_dust_temperatures

Parameters to be renamed (see Issue #607):

• dust_species -> chiaki_dust_model_dust_species
• grain_growth -> chiaki_dust_model_grain_growth
• multi_metals -> chiaki_dust_model_multi_metals
• metal_abundances -> chiaki_dust_model_metal_abundances