Spatio-temporal characterization of Cassiopeia A

Analyzing the X-ray data of supernova remnants (SNRs) are among the most challenging task in the current X-ray astronomy because SNRs are both spatially extended and variable over time. We developed the strategy to track the time-series properties of all the parts constituting a diffuse structure by introducing the free-form image registration technique based on B-spline, and demonstrated the methodology using the Chandra data of Cassiopeia A. We successfully extracted the spatial distribution map of the time variability of continuum luminosity. To our knowledge, this is the first comprehensive characterization of such a dynamic diffuse target both in spatial and temporal viewpoints. We found that each of the four clusters derived by applying k-means algorithm to the extracted light curves has a clear physical meaning distinct from other clusters, which shows that our method is not a mere technique for automation but capable of capturing the underlying physics.