Estimating the column density in molecular clouds with far-infrared and submillimeter emission maps

We have used a numerical simulation of a turbulent cloud to synthesize maps of the thermal emission from dust at a variety of far-IR and submillimeter wavelengths. The average column density and external radiation field in the simulation is well matched to clouds such as Perseus and Ophiuchus. We use pairs of single-wavelength emission maps to derive the dust color temperature and column density, and we compare the derived column densities with the true column density. We demonstrate that longer wavelength emission maps yield less biased estimates of column density than maps made toward the peak of the dust emission spectrum. We compare the scatter in the derived column density with the observed scatter in Perseus and Ophiuchus. We find that while in Perseus all of the observed scatter in the emission-derived versus the extinction-derived column density can be attributed to the flawed assumption of isothermal dust along each line of sight, in Ophiuchus there is additional scatter above what can be explained by the isothermal assumption. Our results imply that variations in dust emission properties within a molecular cloud are not necessarily a major source of uncertainty in column density measurements.