CGV: Small: Quantification and Visualization of Ensemble Uncertainty

不确定性是科学数据所固有的。对不确定性的量化和解释对于全面理解数据至关重要。由于缺乏有效的方法和对现有方法的指导，不确定性表示在高维和多元数据中经常被忽视。该项目系统地研究了气象学应用领域中高维数据的不确定性可视化。这个跨学科的研究团队跨越了可视化、大气科学和统计学等不同领域。研究小组正在研究(1)通过使用自举平均置信区间来量化不确定性，这放宽了对潜在分布的假设；(2)设计了多种不确定性可视化方法，并在集成用户界面中实现；(3)采用定性和定量用户研究对不确定度可视化方法进行评价。得到的工具和方法被部署到环境建模中心。虽然这个项目的主要应用领域是气象学，但学到的原理有可能应用于各种科学领域，包括水文学、地理空间科学、医学成像和生物学。该项目的研究工作被整合到密西西比州立大学的可视化和气象学课程中。

Uncertainty is inherent in scientific data. Quantification and interpretation of the uncertainty are essential to a holistic understanding of the data. Uncertainty representation is often overlooked in high-dimensional and multivariate data due to a lack of effective methods and a lack of guidance on appropriate existing methods. This project systematically investigates uncertainty visualization in high-dimensional data, within the application domain of meteorology. The interdisciplinary research team spans different areas such as visualization, atmospheric science, and statistics. The research team is investigating (1) quantifying uncertainty by using bootstrap mean confidence intervals, which relax assumptions about the underlying distribution; (2) designing a number of uncertainty visualization methods and implement these methods in an integrated user interface; and (3) employing both qualitative and quantitative user studies to evaluate the uncertainty visualization methods.The resulting tools and methods are deployed to the Environmental Modeling Center. Although the primary application area of this project is meteorology, the principles learned have the potential to be applied to a variety of scientific fields, including hydrology, geospatial science, medical imaging, and biology. The research work of this project is integrated into Visualization and Meteorology classes at Mississippi State University.