The modeling of extinction due to dust in the host galaxy in Galacticus is implemented in post-processing, i.e., after the unextincted SED has been predicted. We would have more control over observed colors if we could vary the dust model programmatically, rather than keeping it fixed as in our current scheme. In fact, with @dkorytov we attempted to do this very crudely by varying the "dust factors" in the library, but in the end we abandoned that approach because it gave rise to unwanted to discreteness effects. Based on a recent paper led by Alex Merson and conversation with @abensonca, a promising empirical approach to this problem is to develop a one-parameter family of extinction curves, where the parameter controls the shape of the extinction, and both this parameter and the effective column density are allowed to vary according to galaxy properties such as M* and B/T. The existing Galacticus library, as well as the treatment of dust currently being done by CatSim, would be a good place to look for guidance in building a baseline extinction model with the correct qualitative features. Exploring this alternative dust modeling would be worthwhile to do alongside the planned overhaul of the Galacticus library over the coming year.
The modeling of extinction due to dust in the host galaxy in Galacticus is implemented in post-processing, i.e., after the unextincted SED has been predicted. We would have more control over observed colors if we could vary the dust model programmatically, rather than keeping it fixed as in our current scheme. In fact, with @dkorytov we attempted to do this very crudely by varying the "dust factors" in the library, but in the end we abandoned that approach because it gave rise to unwanted to discreteness effects. Based on a recent paper led by Alex Merson and conversation with @abensonca, a promising empirical approach to this problem is to develop a one-parameter family of extinction curves, where the parameter controls the shape of the extinction, and both this parameter and the effective column density are allowed to vary according to galaxy properties such as M* and B/T. The existing Galacticus library, as well as the treatment of dust currently being done by CatSim, would be a good place to look for guidance in building a baseline extinction model with the correct qualitative features. Exploring this alternative dust modeling would be worthwhile to do alongside the planned overhaul of the Galacticus library over the coming year.
CC @evevkovacs @katrinheitmann @danielsf