Analysis of model output and science data in the Virtual Model Repository (VMR).

Tuesday, 16 December 2014
Darren De Zeeuw, Center for Space Environment Modeling, Ann Arbor, MI, United States and Aaron J Ridley, Univ Michigan, Ann Arbor, MI, United States
Big scientific data not only includes large repositories of data from scientific platforms like satelites and ground observation, but also the vast output of numerical models. The Virtual Model Repository (VMR) provides scientific analysis and visualization tools for a many numerical models of the Earth-Sun system. Individual runs can be analyzed in the VMR and compared to relevant data through relevant metadata, but larger collections of runs can also now be studied and statistics generated on the accuracy and tendancies of model output. The vast model repository at the CCMC with over 1000 simulations of the Earth's magnetosphere was used to look at overall trends in accuracy when compared to satelites such as GOES, Geotail, and Cluster. Methodology for this analysis as well as case studies will be presented.