A study of some current methods of analysing observations of star-forming regions

We present an evaluative study of some current methods utilized in the analysis of infrared (IR) observations of star-forming regions. A series of self-consistent radiative transfer models are constructed, with the outputs analysed using these methods to infer source properties such as the dust temperature, mass, opacity function and density distribution. Any discrepancies between the inferred and model quantities can be attributed to the analysis methods. The range of validity of most methods is smaller than expected, owing to two effects: (i) limited applicability of the Rayleigh-Jeans limit except to very long wavelengths and (ii) significant errors in the isothermal approximation, even when DeltaT(r) < 2 K over 90 per cent of a region. Still, an accurate mean T-dust can be found using a modified Wiens law. This temperature can yield dust masses to within 10-25 per cent - much better than masses inferred from the integrated luminosity. Using long wavelengths (>1000-2000 mum), the opacity index can be determined from the far-IR spectrum to within 20 per cent. Fitting the spectrum yields better results. The density distribution can be somewhat constrained by fitting the surface brightness, for well-resolved sources. Better results are found by fitting the flux spectrum with detailed models.