Abstract: In the process of marine forecasting business, there are many core issues such as multi-source, heterogeneous, massively and strongly correlated marine data storage inefficient, large-scale high-dimensional vector field data visualization slow, multi-forecasting task collaborative interaction heavy and abnormal. These problems have brought great challenges to the comprehensive and multi-angle applications of ocean forecast data. This paper designs and develops a marine forecast monitoring data storage scheme based on the correlation of marine events, a rapid oceanographic forecast data visualization method based on SDR, and a multi-source oceanic forecast data collaborative interaction scheme based on P2P communication. A big data visualization system for ocean forecasting task is established, which provides a fast, dynamic and interactive display effect for multi-source ocean forecast data.

Key words: marine forecast data, high-dimension, strong correlation, data storage, data visualization, collaborative interaction

摘要： 海洋预报业务可视化过程中，存在多源、异构、海量、强关联的海洋数据存储效率低，大规模高维矢量场数据可视化速度慢，多预报任务协同交互负担高，异常多的等核心问题，为海洋预报数据的全方位、多角度应用带来了极大的挑战。设计并研发了基于海洋事件关联度的海洋预报监测数据存储方案、基于SDR的快速海洋预报数据可视化方法、基于P2P通信的多源海洋预报数据协同交互方案，建立了面向海洋预报任务的大数据可视化系统，为多源海洋预报数据提供了快速、动态、易交互的展示效果。

关键词: 海洋预报数据, 超高维, 强关联, 数据存储, 数据可视化, 协同交互