Over the last decade there has been rapid development in our understanding of star formation. This development has primarily been driven by observations of the short timescales on which stars form and the relatively short lifetimes of dense molecular clouds. This has led to a paradigm shift from quasi-static models of star formation to models involving collapse of supersonically turbulent clouds on timescales similar to the dynamical time, spurring the development of hydrodynamic simulations which can form stars out of supercritical clumps within a turbulent cloud. These models have successfully produces self-gravitating protostars with the mass distributions that resemble the observed IMF, however it is difficult to compare these simulations directly to millimeter and submillimeter observations of rotovibrational molecular transitions within star-forming clouds. Unfortunately this results in a gap between theory and observation. We propose stepping into this gap and creating a comprehensive radiative transfer platform which can be used to simulate the molecular line emission from a wide variety of simulations in a wide selection of molecular transitions.

The goal of this project is to develop a 3-D Monte Carlo radiative transfer simulation on a heterogeneous computational cluster using a distributed memory space system. We call this platform the Molecular Virtual Observatory as it is intended primarily to simulate rotovibrational molecular transitions. This will allow for a direct statistical comparison of 2-D and 3-D hydrodynamic simulations with observations of star-forming molecular clouds. Although Monte Carlo radiative transfer is fairly well understood, the consistent problem has been that it tends to converge very slowly, however Monte Carlo techniques are also inherently parallelizable, allowing one to easily split up the problem across multiple processors or computers. We are developing this code to run on an existing cluster of 24 computers available at our institution.

Once the Molecular Virtual Observatory is complete we can begin producing synthetic observations of existing hydrodynamic simulations. These simulations will be made available by UC San Diego's Computational Analysis and Data Center. Our synthetic observations can be statistically compared to millimeter and submillimeter observations of similar regions using techniques such as principle component analysis or the spectral correlation function in order to ascertain weather the model accurately simulates the physical and chemical state and evolution of the star-forming region. This will become especially important as the Atacama Large Millimeter Array (ALMA) allows us to probe the physical and chemical structure of gas disks around young stars from which planets form. By creating synthetic observations of these regions the Molecular Virtual Observatory will play a key role in testing our understanding of the earliest stages of planet formation.