Properties of Molecular Clouds

Plotted here are dust temperature (top) and column density (bottom) maps of the Perseus molecular cloud. These maps were made using data from Spitzer (at 70 and 160 microns), IRAS (at 100 microns) and 2MASS (at J, H and K band). Taken from Schnee et al. (2008).

Nearby molecular clouds such as Perseus offer astronomers regions to study star formation in both clustered and isolated environments. In the image above you can see that Perseus is characterized by filaments of high column density and low temperature, with two clusters of forming stars (IC348 and NGC1333) where the pixels are saturated. The starless cores and protostars in Perseus are found almost exclusively in the dark filaments and clusters.

My research on molecular clouds has largely focused on the dust properties and mapping of cloud structure. As part of my thesis, I mapped the dust temperature and column density in the Perseus, Ophiuchus and Serpens molecular clouds. I used the ratio of the 60 and 100 micron IRAS maps to derive dust temperature, and I calibrated the far-infrared optical depth with the near-infrared (2MASS) reddening to estimate the column density. As part of this calibration I found that the dust emission/absorption properties of the dust vary from region to region within a molecular cloud and between the different molecular clouds. To see the paper describing these results in greater detail, click here.

By using Spitzer maps of Perseus at 70 and 160 microns, I showed that it is possible to create much higher resolution images of a nearby molecular cloud while at the same time correcting for variations in the dust emissivity and emission mechanism. The dust properties in Perseus vary with column density and temperature in a manner consistent with grain growth. To see the paper describing these results, including the image shown at the top of this page, click here.

Two of the primary methods for deriving column density in a molecular cloud are to use dust emission in the far-infrared and dust absorption in the near-infrared of the light from stars behind the cloud. In a comparison of the two column density estimates, I have found that the scatter is on the order of a few tens of percent. As part of my thesis, and working with Tom Bethell, we used a numerical simulation of a turbulent molecular cloud with radiative transfer to quantify the error in the emission-derived column density caused by line of sight variations in the dust temperature. It turns out that at the far-infrared wavelengths measured by IRAS, the incorrect assumption of isothermality creates an error in the emission-derived column density comparable to that observed in Perseus. The errors are greater at higher column densities, but can be diminished by observing at longer wavelengths. Click here to see our paper.