In climate models, the physical domains of the atmosphere and of the ocean are represented by three-dimensional computational grids. For each cell and at regular time intervals, the derived values of various physical quantities need to be stored in order to later analyse and visualise the results.The finer the computational grid is, the more realistically important phenomena can be simulated. Typical data sets with results of climate models are therefore three-dimensional, multivariate and time-dependent. Each simulation of the new high resolution weather- and climate models produce hundreds of Terrabytes of binary data. Turning this kind of data into visual information that captures the core information and enables an intuitive analysis is quite challenging.

ESiWACE takes a multi-tiered approach by expanding the capabilities of the high performance data analysis system Ophidia at CMCC (see upper right image), and extending the 3D visualisation software ParaView (see lower right image) for use with the different models supported by ESiWACE. We also pave the way for in-situ analysis by evaluating the use of ParaView Catalyst in combination with the ICON model at DKRZ.