基于雷达图序列的海洋多维数据可视化方法

Ocean Hydrological Data Visualization Based on Radar Chart Series

摘要: 传统的海洋水文数据可视化方法通常只对单一要素进行显示，而多维数据可视化方法通常只在二维平面坐标系上显示多要素数据。本文针对海洋高维多元数据提出了一种通过添加时空维度坐标轴，将雷达图扩展到三维空间的方法，绘制由二维雷达图和一个时空维构成的雷达图序列Radar Chart Series，融合了剖面图和雷达图各自的信息，既在每个雷达图表示多个水文要素的数据信息，又在时空维度上反映各要素沿某一时空维度的变化趋势，从而在全局上展现整个数据集分布特征以及各个研究对象(要素)之间可能存在的关系。实验结果表明，雷达图序列方法在很大程度上拓宽了一个图所能承载的数据信息，对数据中的隐藏信息进行了更深层次的挖掘。

Abstract: Traditional data visualization can only visualize single element of ocean hydrological data. Multidimensional data visualization can visualize multi-element in planar coordinate. During to its short-age in displaying multidimensional multivariate data, a new method to add dimension was pro-posed, which could extend Radar Chart into 3D space (Radar Chart Series). So that multidimensional multivariate data could be displayed as polygonal plane cluster for marine elements showing and analyzing. Results reveal that Radar Chart Series can effectively analyze characters and provide a powerful tool to reveal the rules of abstract data and the complex oceanic phenomena.