Simulation modeling has become an important component of the modeling process for many scientific phenomena, particularly when physical measurements are difficult or expensive to collect. As increasingly complex problems are explored, large amounts of data are being generated using large scale simulations. Analysis and visualization methods are needed to extract relevant information from these sources of Big Data. In many cases, traditional analysis and visualization methods have limitations that are exceeded by today's massive datasets. An integrated approach to analysis and visualization that combines the rigor of statistical methods with the power of computation and visualization is needed to address a variety of analytical challenges including change detection, feature identification, outlier detection, and data reduction. In addition to traditional post-processing of simulation outputs, analysis of large scale simulations may require in situ analysis that can be performed while a simulation is running or on-the-fly analysis techniques that can be implemented within state-of-the-art visualization systems.