We introduce ProSpect, a generative galaxy spectral energy distribution code that can also be used for parameter inference. ProSpect comes with two popular families of stellar population libraries (BC03 and EMILES), and a large variety of methods to construct star formation and metallicity histories. It models dust through the use of a Charlot & Fall (2000) attenuation model with re-emission using Dale, et al. (2014) far-infrared templates. It also has the ability to model AGN through the inclusion of a simple AGN and hot torus model. Finally, it makes use of MAPPINGS-III photoionisation tables to produce line emission features. We test the generative and inversion utility of ProSpect through application to the Shark semi-analytic code, and informed by these results produce fits to the final photometric catalogues produces by the Galaxy and Mass Assembly Survey (GAMA). As part of the testing of ProSpect, we also produce a range of simple photometric stellar mass approximations covering a range of filters for both observed frame and rest frame photometry.This work is interested in combining the best methods, codes, tables, philosophies, practices and practicalities to produce software (ProSpect) that whilst known to be imperfect, is at least useful, is simple to use, and rapid to run. The primary use case for ProSpect is to produce useful and self consistent UV-FIR outputs for current and future semianalytic models , however due to its structure it is deliberately simple to execute ProSpect in a Bayesian generative mode that allows inversion of interesting galaxy properties. The most effort in this regard

We present catalogues of stellar masses, star formation rates, and ancillary stellar population parameters for galaxies spanning 0 < z < 9 from the Deep Extragalactic VIsible Legacy Survey (DEVILS). DEVILS is a deep spectroscopic redshift survey with very high completeness, covering several premier deep fields including COSMOS (D10). Our stellar mass and star formation rate estimates are self-consistently derived using the spectral energy distribution (SED) modelling code ProSpect, using well-motivated parameterisations for dust attenuation, star formation histories, and metallicity evolution. We show how these improvements, and especially our physically motivated assumptions about metallicity evolution, have an appreciable systematic effect on the inferred stellar masses, at the level of ∼ 0.2 dex. To illustrate the scientific value of these data, we map the evolving galaxy stellar mass function (SMF) and the SFR-M⋆ relation for 0 < z < 4.25. In agreement with past studies, we find that most of the evolution in the SMF is driven by the characteristic density parameter, with little evolution in the characteristic mass and low-mass slopes. Where the SFR-M⋆ relation is indistinguishable from a power-law at z > 2.6, we see evidence of a bend in the relation at low redshifts (z < 0.45). This suggests evolution in both the normalisation and shape of the SFR-M⋆ relation since cosmic noon. It is significant that we only clearly see this bend when combining our new DEVILS measurements with consistently derived values for lower redshift galaxies from the Galaxy And Mass Assembly (GAMA) survey: this shows the power of having consistent treatment for galaxies at all redshifts.

We combine the Shark semi-analytic model of galaxy formation with the ProSpect software tool for spectral energy distribution (SED) generation to study the multiwavelength emission of galaxies from the far-ultraviolet (FUV) to the far-infrared (FIR) at 0 ≤ z ≤ 10. We produce a physical model for the attenuation of galaxies across cosmic time by combining a local Universe empirical relation to compute the dust mass of galaxies from their gas metallicity and mass, attenuation curves derived from radiative transfer calculations of galaxies in the EAGLE hydrodynamic simulation suite, and the properties of Shark galaxies. We are able to produce a wide range of galaxies, from the z = 8 star-forming galaxies with almost no extinction, z = 2 submillimeter galaxies, down to the normal star-forming and red sequence galaxies at z = 0. Quantitatively, we find that Shark reproduces the observed (i) the z = 0 FUV-to-FIR, (ii) 0 ≤ z ≤ 3 rest-frame K-band, and (iii) 0 ≤ z ≤ 10 rest-frame FUV luminosity functions, (iv) z ≤ 8 UV slopes, (v) the FUV-to-FIR number counts (including the widely disputed 850µm), (vi) redshift distribution of bright 850µm galaxies and (vii) the integrated cosmic SED from z = 0 to z = 1 to an unprecedented level. This is achieved without the need to invoke changes in the stellar initial mass function, dust-to-metal mass ratio, or metal enrichment timescales. Our model predicts star formation in galaxy disks to dominate in the FUV-to-optical, while bulges dominate at the NIR at all redshifts. The FIR sees a strong evolution in which disks dominate at z ≤ 1 and starbursts (triggered by both galaxy mergers and disk instabilities, in an even mix) dominate at higher redshifts, even out to z = 10.