|
The abundance and diversity of data provided by the recent large, multi-wavelength, surveys of star-forming regions (e.g. COMPLETE, c2d) allows us to measure what should be "the same" physical quantities in a variety of different ways. For example, one can measure column density: using traditional methods like spectral line emission from gas, thermal emission from dust, or extinction due to dust; or by using newer methods like analyzing the scattering of the ISRF off of dust ("Cloudshine"). If our assumptions about various ratios (e.g. gas-to-dust, molecular abundances, etc.) are correct, then various techniques for measuring similar quantities should give the same answer. Recent work, including our own, has shown that there are many surprisingly large inconsistencies amongst not only techniques for measuring column density, but also amongst the structure-finding algorithms (e.g. CLUMPFIND) that are used to catalog and categorize structures. In this talk I will show some dramatic new illustrations of how views of cloud structure can, and do, differ due to biases and uncertainties inherent in observing techniques. I will discuss biases and uncertainties caused by the effects of: 1) non-uniform line-of-sight temperature distribution on column density estimates; 2) dust properties on inter-comparisons of extinction and thermal emission; 3) threshold and topology choices in structure finding algorithms; 4) wavelength choices in searches for young stars in color-color diagrams; and 5) chemical and opacity effects on (particularly 13CO) spectral-line maps. Looking toward the future, I will also try to offer the best prescriptive solution possible to overcoming each observing bias. The talk will intentionally draw upon many different kinds of observations to be presented at this meeting, and it should offer a good forum for discussion of how they should best be combined in the future to enhance our understanding of the early phases of star formation.
|
