Information visualization tool for large-scale scientific data analysis

Miyamura, Hiroko  ; Kawamura, Takuma ; Suzuki, Yoshio   ; Idomura, Yasuhiro   ; Takemiya, Hiroshi

In this paper, we propose an information visualization tool for the analysis of multivariate time series simulation datasets. For the implementation of this tool, we use a spatio-temporal map. The spatio-temporal map provides a method for mapping the distribution of physical data in a two-dimensional visualization space for space and time axis. It is possible to use parallel coordinate visualization technique of multivariate dataset for analysis of time series and multivariate datasets. In this tool, the physical data to be observed are first selected from the parallel coordinates, then, the spatial-temporal distribution of the selected physical data is observed using a spatio-temporal map. We also describe the effects of this tool by applying the simulation results of earthquake response of a nuclear test reactor.