Automated supershell recognition in spiral galaxies employing hydrodynamic simulations

The search for expanding supershells in nearby galaxies has been the focus of substantial observational effort since the discovery of huge HI bubbles in the Magellanic clouds. Until recently, studies of this nature were limited: initially due to the lack of detailed models describing the predicted appearance of supershells, subsequently because it is challenging to quantify the similarity of model and data over numerous channel maps by visual inspection. Thilker et al. (\cite{Thilker2}) first established a method for automated identification and classification of supershells in spiral galaxies. Their object recognition technique is capable of locating complex 3-dimensional patterns in a spectral-line datacube via cross-correlation analysis. We report on various refinements to the basic method and discuss the implications for one particular galaxy (NGC 2403). Most significantly, we now incorporate realistic hydrodynamic models of superbubble structure. We also introduce more robust treatment of shell significance, especially for the case of overlapping structures.