slope-attenuation relation

[changhoonhahn]

Observations find a relationship between the slope and amplitude of dust attenuation in star-forming galaxies: galaxies with high attenuation have shallower slopes. For example in Salim & Narayanan (2020) they present the following relation for star-forming galaxies:

Current radiative transfer models are able to more-or-less reproduce this slope-attenuation relation: e.g. Chevellard+(2013) compared to Salim+(2018)

Trayford+(2020) compared to Salim+(2018) (red scatter points vs blue shaded)

Our dust model also more-or-less reproduce the slope-attenuation relation for star-forming galaxies: Yet overall, our models find an overall steeper slope, especially at high attenuation.

Examining where different subpopulations of galaxies lie on the slope-attenuation relation sheds some light: Blue and Green are star-forming. Blue has logSSFR > -9.75; Green has -9.75 > logSSFR > -10.5. Markers with black edges have M_* > 10^10.5Msun. (ignore the contours)

Lower mass star-forming galaxies have significantly shallower slopes. Our overall steeper slope is likely due to the fact that our sample has a high stellar mass limit.

Here's how the slope-attenuation relation looks for the entire galaxy population (including quiescent galaxies): Orange (low SSFR), red (lower SSFR), purple (lowest SSFR) are all quiescent galaxies.

There's a couple of interesting points in these figures:

• [x] _Why do star-forming galaxies in EAGLE and TNG have such different AV ranges? TNG SF galaxies have A_V > 0.75 while most EAGLE SF galaxies have A_V < 0.75.
• [x] Why do quiescent galaxies in TNG have the opposite slope-attenuation relation?

[changhoonhahn]

Latest updates to the slope-attenuation results from the updated SFR=0 prescription*

slope-attenuation relation of star-forming galaxies

slope-attenuation relation of galaxy subpopulations

• TNG has a higher A_V range than EAGLE because TNG predicts intrinsically bluer galaxy populations.

• With the updated SFR=0 prescription, we no longer see the unusual trend in quiescent TNG galaxies.

*The new prescription assigns SFR = SFR_min to SFR=0 galaxies

[changhoonhahn]

The fact that our dust model reproduces the attenuation-slope relation is neat since the relation is a result of dust scattering dominating at low attenuation and dust absorption dominating at high attenuation. We're able to infer the relation purely from the observations with a model that has no priors on it. This is a good demonstration of the forward modeling approach and its advantages. Other than that, I don't think we learn much from this comparison so I'm going to simplify the figure to: