Shielding

The radiation field defined through the UV_field parameter represents the radiation that is incident on the outside of the gas cloud. However, we also need to consider the self-shielding of the gas cloud, which attenuates the incident radiation field. We do this by specifying a shielding length, L_sh, which characterises the size of the cloud. This then defines the total hydrogen column density that attenuates the radiation field as N_Htot = n_Htot * L_sh, where n_Htot is the total number density of hydrogen nuclei, and the column densities of individual species, e.g. N_HI = n_HI * L_sh etc., which depend on the current abundances of those species. The cross sections data files used in CHIMES then tabulate how the various photoionisation and photoheating rates are attenuated as a function of these column densities, which accounts for how the shape of the UV spectra changes with column density (i.e. spectral hardening) as well as the overall reduction in the strength of the radiation field. See Section 3 of Richings et al. (2014b) for details.

In CHIMES Driver, we use the shield_mode parameter to specify how to calculate the shielding length that is then passed to the chemistry solver. The possible options are described below.

shield_mode Description
None
Shielding is disabled, and the optically thin rates are used.
read-in
The shielding column density of each gas particle is read in from the
snapshot. This needs to be given as the total hydrogen column density,
N_Htot, in units of cm^-2. The shielding length is then calculated as
L_sh = N_Htot / n_Htot, where n_Htot is the total number density
of hydrogen nuclei. This option can only be used if
IO_mode == snapshot.
Jeans
The shielding length of each particle, or each point in the grid, is set
to the Jeans length, as a function of gas temperature and density. This
is multiplied by the shield_length_factor parameter, and is limited
according to the max_shield_length parameter (in cm).
Colibre
The shielding length of each particle, or each point in the grid, is set
according to the COLIBRE model (based on Ploeckinger & Schaye 2020,
albeit with some modifications).