Historically, almost all our information about the current and past star-formation properties of normal galaxies has been based upon spatially integrated measurements in the ultraviolet (UV), visible and near-infrared (NIR) spectral regimes. However, star-forming galaxies contain dust which absorbs some fraction of the emitted starlight, re-radiating it predominantly in the far-infrared (FIR)/sub-millimeter (submm) range.
In view of this, the measurement of current and past star-formation in galaxies - and indeed of the universe as a whole - requires a quantitative understanding of the role different stellar populations play in powering the FIR/submm emission. For this both optical and FIR/submm data need to be used, as they contain complementary information about the distribution of stars and dust. The problem is very complex. Images of galaxies taken in the stellar light and dust emission show that galaxies are very inhomogeneous systems, presenting both small scale and diffuse components, both in the UV and in the FIR. For instance small scale structures show that part of the stellar light can be locally absorbed and re-radiated in the FIR. The diffuse emission shows that another part of the stellar photons can travel some distance in the disk before being scattered or absorbed. In order to quantitatively model these processes one needs to make a self-consistent calculation of the transport of radiation and its re-emission in galaxy disks. The challenge is to identify a model which is sufficiently simple to make the problem tractable, but which can still predict the essential elements of observed SEDs and morphologies. Here we review progress made towards identifying such models for normal galaxies. Models for predicting SEDs for starburst galaxies are considered in the review of Dopita .