MAGIC: Marching Cubes Isosurface Uncertainty Visualization for Gaussian Uncertain Data With Spatial Correlation.

Area of Science:

• Scientific Visualization
• Uncertainty Quantification
• Computational Geometry

Background:

• Isosurface visualization of uncertain data requires accounting for spatial correlations to avoid errors.
• Existing methods for correlated uncertain data lack analytical formulations and rely on computationally expensive Monte Carlo sampling.
• Prior treatments of isosurface uncertainty with spatial data correlation have significant limitations.

Purpose of the Study:

• To develop an efficient, closed-form analytical framework for quantifying uncertainty in isosurfaces generated by the Marching Cubes algorithm.
• To address the lack of analytical solutions for Gaussian uncertain data with spatial correlation in isosurface visualization.
• To provide a computationally efficient and accurate method for uncertainty quantification in correlated uncertain data.

Main Methods:

• Leveraged Hinkley's derivation on the ratio of Gaussian distributions to create closed-form solutions.
• Developed the Marching Cubes algorithm for Gaussian uncertain data with spatial correlation (MAGIC) framework.
• Utilized many-core processors to accelerate the analytical solutions.

Main Results:

• Achieved significant speed-up and enhanced accuracy in uncertainty quantification compared to Monte Carlo methods.
• Demonstrated speed-ups up to 585x through many-core processor acceleration.
• Validated the correlation-aware uncertainty framework on meteorology, urban flow, and astrophysics datasets.

Conclusions:

• The proposed closed-form framework (MAGIC) efficiently quantifies uncertainty in isosurfaces for correlated Gaussian data.
• The analytical approach overcomes the limitations of Monte Carlo methods, offering improved accuracy and speed.
• The framework is integrable with production visualization tools, enabling broader impact in scientific visualization.