Due to mass and temporal resolutions of the simulations, a handful of galaxies have SFR=0 (see notebook). The dust model, which depends on log SSFR, cannot be directly used for these galaxies.
At the moment, I neglect these galaxies using an ad-hoc prescription for directly sampling their observables from the SDSS distribution of quiescent galaxies --- i.e. directly sampling M_r, G-R, FUV-NUV from SDSS quiescent galaxies. However, this is not ideal since this
ignores the SFR dependence of quiescent galaxies (SFR=0 galaxies have lower SFR than SFR > 0 quiescent galaxies)
affects the observable distribution of other quiescent galaxies.
Here's what happens when I sample the SFR between the range [0, SFR_min] for SFR=0 galaxies:
Due to mass and temporal resolutions of the simulations, a handful of galaxies have SFR=0 (see notebook). The dust model, which depends on log SSFR, cannot be directly used for these galaxies.
At the moment, I neglect these galaxies using an ad-hoc prescription for directly sampling their observables from the SDSS distribution of quiescent galaxies --- i.e. directly sampling M_r, G-R, FUV-NUV from SDSS quiescent galaxies. However, this is not ideal since this
Here's what happens when I sample the SFR between the range [0, SFR_min] for SFR=0 galaxies: