Making Sense of Scientific Simulation Ensembles With Semantic Interaction

Abstract

In the study of complex physical systems, scientists use simulations to study the effects of different models and parameters. Seeking to understand the influence and relationships among multiple dimensions, they typically run many simulations and vary the initial conditions in what are known as ‘ensembles’. Ensembles are then a number of runs that are each multi‐dimensional and multi‐variate. In order to understand the connections between simulation parameters and patterns in the output data, we have been developing an approach to the visual analysis of scientific data that merges human expertise and intuition with machine learning and statistics. Our approach is manifested in a new visualization tool, GLEE (Graphically‐Linked Ensemble Explorer), that allows scientists to explore, search, filter and make sense of their ensembles. GLEE uses visualization and semantic interaction (SI) techniques to enable scientists to find similarities and differences between runs, find correlation(s) between different parameters and explore relations and correlations across and between different runs and parameters. Our approach supports scientists in selecting interesting subsets of runs in order to investigate and summarize the factors and statistics that show variations and consistencies across different runs. In this paper, we evaluate our tool with experts to understand its strengths and weaknesses for optimization and inverse problems.

Improving the visual communication of environmental model projections

“…As previously mentioned, numerous studies have also linked familiarity with visualisation preferences, particularly in terms of visual appeal (e.g. 42 , 54 , 55 ). Despite this, the final rankings of the visualisations did not follow the ordering of the visualisations that were most familiar to the survey participants.…”

DeepCinema: Adding Depth with X3D Image-Based Rendering

“…One exciting future application will be the integration of depthenhanced IBR objects to Immersive Analytic Workspaces [Dwyer et al 2016] [Marriott et al 2018] [Skarbez et al 2019], such as GLEE [Dahshan et al 2019], where Cinema thumbnails are used to drive human-in-the-loop Machine Learning known as Semantic Interaction [Endert et al 2012] [Dowling et al 2018]. Semantic Interaction in 2D workspaces has shown benefits, especially in the visualization of ensembles.…”

Visual Analysis and Detection of Contrails in Aircraft Engine Simulations

“…Coordinated multiple-views have been adopted to analyze both input parameters and output ensemble data. These aggregated views may include multi-chart visualization [18,19,39], colored overlays [9], series of parallel coordinates plots [49], or various types of tracking graphs [13,52,94]. Luciani et al [50] used multiple-linked views to explore multi-run ensemble simulation and facilitated the understanding of ensemble characteristics, but did not consider the correlation between I/O parameters or the direct comparison between ensemble members.…”