We present the first radiative transfer (RT) model of a non-edge-on disk galaxy in which the large-scale geometry of stars and dust is self-consistently derived through fitting of multiwavelength imaging observations from the UV to the submm. To this end we used the axi-symmetric RT model of Popescu et al. and a new methodology for deriving geometrical parameters, and applied this to decode the{spectral energy distribution (SED) of M33. We successfully account for both the spatial and spectral energy distribution, with residuals typically within $7\%$ in the profiles of surface brightness and within $8\%$ in the spatially-integrated SED. We predict well the energy balance between absorption and re-emission by dust, with no need to invoke modified grain properties, and we find no submm emission that is in excess of our model predictions. We calculate that $80\pm8\%$ of the dust heating is powered by the young stellar populations. We identify several morphological components in M33, a nuclear, an inner, a main and an outer disc, showing a monotonic trend in decreasing star-formation surface-density ($\Sigma_{\rm SFR}$) from the nuclear to the outer disc. In relation to surface density of stellar mass, the $\Sigma_{\rm SFR}$ of these components define a steeper relation than the "main sequence" of star-forming galaxies, which we call a "structurally resolved main sequence". Either environmental or stellar feedback mechanisms could explain the slope of the newly defined sequence. We find the star-formation rate to be ${\rm SFR}=0.28^{+0.02}_{-0.01}{\rm M}_{\odot}{\rm yr}^{-1}$.

We present an extensively updated version of the purely ray-tracing 3D dust radiation transfer code DART-Ray. The new version includes five major upgrades : 1) a series of optimizations for the ray-angular density and the scattered radiation source function; 2) the implementation of several data and task parallelizations using hybrid MPI+OpenMP schemes; 3) the inclusion of dust self-heating; 4) the ability to produce surface brightness maps for observers within the models in HEALPix format; 5) the possibility to set the expected numerical accuracy already at the start of the calculation. We tested the updated code with benchmark models where the dust self-heating is not negligible. Furthermore, we performed a study of the extent of the source influence volumes, using galaxy models, which are critical in determining the efficiency of the DART-Ray algorithm. The new code is publicly available, documented for both users and developers, and accompanied by several programmes to create input grids for different model geometries and to import the results of N-body and SPH simulations. These programmes can be easily adapted to different input geometries, and for different dust models or stellar emission libraries.

Low luminosity galaxies may be the building blocks of more luminous systems. Southern African Large Telescope (SALT) observations of the low luminosity, early-type galaxy NGC59 are obtained and analysed. These data are used to measure the stellar population parameters in the centre and off-centre regions of this galaxy, in order to uncover its likely star formation history. We find evidence of older stars, in addition to young stars in the emission line regions. The metallicity of the stellar population is constrained to be [Z/H] ~ -1.1 to -1.6, which is extremely low, even for this low luminosity galaxy, since it is not classed as a dwarf spheroidal galaxy. The measured [alpha/Fe] ratio is sub-solar, which indicates an extended star formation history in NGC59. If such objects formed the building blocks of more massive, early-type galaxies, then they must have been gaseous mergers, rather than dry mergers, in order to increase the metals to observed levels in luminous, early-type galaxies.