Introductory overview: Recommendations for approaching scientific visualization with large environmental datasets
Introductory overview: Recommendations for approaching scientific visualization with large environmental datasets
复制标题
DOI：
10.1016/j.envsoft.2021.105113
复制
发表时间：
2021-06
期刊：
影响因子：
--
通讯作者：
C. Kelleher;A. Braswell
中科院分区：
文献类型：
--
作者：
C. Kelleher;A. Braswell
科学可视化是向各种受众传达结果和发现的基础。随着新的大型环境数据集的创建不断增加，这就需要新的方案和建议来创建有效的可视化。在这篇综述中，我们回顾了科学可视化的基础，以及在大数据的四个V（体积，多样性，准确性和速度）的背景下对大型数据集进行可视化的考虑。使用大数据集需要决定是否聚合或保留细节，分组以进行比较的方法，以及考虑如何在多维空间中最好地显示复杂数据。为了实现更有效的可视化，我们提供了一些关于可视化过程中面临的常见决策的注意事项。这些建议附有应用于现有大型数据集的示例。虽然我们的建议只是这样，但它们鼓励在可视化科学数据集时对所面临的选择的意向性和意识。
Scientific visualizations are the foundation for communicating results and findings to a variety of audiences. As the creation of novel and large environmental datasets has grown, this has necessitated new schemes and recommendations for creating effective visualizations. In this overview, we review the foundations of scientific visualization and considerations for visualization of large datasets within the context of the four Vs of big data (volume, variety, veracity, and velocity). Using big datasets requires making decisions as to whether to aggregate or preserve details, approaches for grouping to enable comparisons, and considering how best to show complex data in many-dimensional space. To enable more effective visualizations, we provide several considerations regarding common decisions faced during the visualization process. These recommendations are accompanied by examples applied to existing large datasets. While our recommendations are just that, they encourage intentionality and awareness of the choices faced when visualizing scientific datasets.