2013A&A...550A..29P

The question how the initial conditions in a star-forming region affect the resulting mass function of the forming stars is one of the most fundamental open topics in star formation theory. We want to characterize the properties of the cold dust clumps in the Carina nebula complex, which is one of the most massive star forming regions in our Galaxy and shows a very high level of massive star feedback. We derive the clump mass function (ClMF), explore the reliability of different clump extraction algorithms, and investigate the influence of the temperatures within the clouds on the resulting shape of the ClMF. We analyze a 1.25°x1.25° wide-field submillimeter map obtained with LABOCA at the APEX telescope, which provides the first spatially complete survey of the clouds in the Carina nebula complex. We use the three clump-finding algorithms CLUMPFIND, GAUSSCLUMPS and SExtractor to identify individual clumps and determine their total fluxes. In addition to assuming a common ``typical'' temperature for all clouds, we also employ an empirical relation between cloud column densities and temperature to determine an estimate of the individual clump temperatures, and use this to determine individual clump masses. We find that the ClMFs resulting from the different extraction methods show considerable differences in their shape. While the ClMF based on the CLUMPFIND extraction is very well described by a power-law (for clump masses well above the completeness limit), the ClMFs based on the extractions with GAUSSCLUMPS and SExtractor are better represented by a log-normal distribution. We also find that the use of individual clump temperatures leads to a shallower ClMF slope than the (often used) assumption of a common temperature (e.g. 20K) of all clumps. The power-law of dN/dM ∝M^–1.95 we find for the CLUMPFIND sample is in good agreement with ClMF slopes found in previous studies of the ClMFs of other regions. The dependence of the ClMF shape (power-law versus log-normal distribution) on the employed extraction method suggests that observational determinations of the ClMF shape yields only very limited information about the true structure of the cloud. Interpretations of log-normal ClMF shape as a signature of turbulent pre-stellar clouds versus power-law ClMFs as a signature of star-forming clouds may be taken with caution for a single extraction algorithm without additional information.