Previous studies of the sizes and central densities of simulated massive, star-forming galaxies have shown that, without AGN feedback, they become too compact relative to observed galaxies at 𝑧 2. In this paper, we perform high-resolution re-simulations of a massive (𝑀 ★ ∼ 10 11 M ) galaxy at 𝑧 ≈ 2, drawn from the Feedback in Realistic Environments (FIRE) project. In the fiducial simulation, which does not include AGN feedback, the galaxy experiences a rapid starburst and shrinking of its half-mass radius at 𝑧 ≈ 2.3. In this paper, we experiment with driving mechanical AGN winds at specified, constant rates, using a state-of-the-art hyper-Lagrangian refinement technique to increase particle resolution. These winds reduce the gas surface density in the inner regions of the galaxy, suppressing the very compact starburst and maintaining a half-mass radius that is approximately constant with time. We run radiative transfer calculations to study the observable impact of AGN feedback on multi-wavelength continuum emission and find notable differences between the simulations with and without winds, in both the magnitude and spatial extent of observable short and long-wavelength flux. When AGN winds are included, the suppression of the compact, dusty starburst results in lowered flux at FIR wavelengths (due to decreased star formation) but increased flux at optical-to-near-IR wavelengths (due to decreased dust attenuation, in spite of the lowered star formation rate), relative to the case without AGN winds. The FIR half-light radius decreases from ∼ 1 kpc to ∼ 0.1 kpc in < 40 Myr when AGN winds are not included, but increases to ∼ 2 kpc when they are. The impact of AGN-driven winds at shorter wavelengths is less intuitive, with half-light radii of optical-NIR emission remaining approximately constant over 𝑡 = 35 Myr, for simulations with and without AGN winds. In the case without winds, this occurs despite the rapid compaction, and is due to heavy dust obscuration in the inner regions of the galaxy. This work highlights the importance of forward-modelling when comparing simulated and observed galaxy populations.

Single flux density measurements at observed-frame sub-millimeter and millimeter wavelengths are commonly used to probe dust and gas masses in galaxies. In this Letter, we explore the robustness of this method to infer dust mass, focusing on quiescent galaxies, using a series of controlled experiments on four massive haloes from the Feedback in Realistic Environments (FIRE) project. Our starting point is four star-forming, central galaxies at seven redshifts between 𝑧 = 1.5 and 𝑧 = 4.5. We generate modified quiescent galaxies that have been quenched for 100 Myr, 500 Myr, or 1 Gyr prior to each of the studied redshifts by re-assigning stellar ages. We derive spectral energy distributions for each fiducial and modified galaxy using radiative transfer. We demonstrate that the dust mass inferred is highly dependent on the assumed dust temperature, 𝑇 dust , which is often unconstrained observationally. Motivated by recent work on quiescent galaxies that assumed 𝑇 dust ∼ 25 K, we show that the ratio between dust mass and 1.3 mm flux density can be higher than inferred by up to an order of magnitude, due to the considerably lower dust temperatures seen in non star-forming galaxies. This can lead to an underestimation of dust mass (and, when sub-mm flux density is used as a proxy for molecular gas content, gas mass). This underestimation is most severe at higher redshifts, where the observed-frame 1.3 mm flux density probes rest-frame wavelengths far from the Rayleigh-Jeans regime, and hence depends super-linearly on dust temperature. We fit relations between ratios of rest-frame far-infrared flux densities and mass-weighted dust temperature that can be used to constrain dust temperatures from observations and hence derive more reliable dust and molecular gas masses.