eFESTA: Ensemble Feature Exploration with Surface Density Estimates.

Area of Science:

• Scientific visualization
• Computational geometry
• Data analysis

Background:

• Analyzing complex 3D ensemble simulation data presents challenges in identifying dominant spatial patterns.
• Existing methods often rely on field datasets, limiting their applicability to surface feature data.

Purpose of the Study:

• To propose a novel method for modeling the spatial distribution of 3D surface features.
• To develop a technique for extracting and visualizing major trends in ensemble surface features.
• To enable analysis without direct reliance on scalar or vector field data.

Main Methods:

• Surface Density Estimate (SDE): A kernel density estimation applied to surface features represented as polygon meshes.
• Bandwidth Selection Algorithm: An automated approach for choosing an appropriate kernel bandwidth for SDE.
• ensemble Feature Exploration based on Surface densiTy EstimAtes (eFESTA): A hierarchical method for organizing and exploring density fields derived from SDE.

Main Results:

• SDE effectively models spatial distributions using only surface feature geometry (polygon meshes).
• eFESTA successfully extracts and visualizes dominant trends in ensemble surface features.
• Demonstrated applications include isosurfaces, Lagrangian coherent structures, and streamsurfaces in various flow simulations.

Conclusions:

• The proposed SDE and eFESTA provide a robust framework for analyzing and visualizing ensemble surface features.
• This approach offers a valuable alternative when field datasets are unavailable or computationally expensive.
• The method enhances the understanding of complex spatial patterns in scientific simulations.