海量数据驱动下的高分辨率海洋数值模式关键算法研究

海洋数值模式是定量描述海洋物理现象及其变化的数值模型，也是海洋与气候研究、预测的核心工具。目前，海洋数值模式正朝着高分辨率和多物理过程的方向发展。随着分辨率的提高、物理过程的细化，模式不确定物理参数的合理估计、大规模并行计算的性能以及快速准确的后处理与可视化已成为其发展与应用的瓶颈。高性能计算机和深度学习等技术的快速发展，为海洋数值模式的突破提供了机遇与挑战。海浪和海洋环流是海洋中最重要的物理现象。本项目将针对海浪模式和海洋环流模式在发展中遇到的关键瓶颈问题，基于国产E级超算平台和海量海洋数据，开展大规模高效并行算法研究，提高模式计算速度；基于深度学习的新方法开展模式物理参数合理估计，提高模式模拟能力；突破海量数据下可视化关键算法，提高模式后处理和可视化应用水平，最终提升我国海洋模式发展和应用水平，促进海洋科学、高性能计算以及人工智能的深入交叉融合，形成基于国产神威E级系统的大型示范应用。

The ocean model is the key tool for ocean and climate research and prediction, which is a numerical model for the quantitative description of marine phenomena based on the physical laws. Now, the main stream of ocean model development focuses on higher resolution and more accurate parameterization of unresolved physical processes. With the higher resolution and the more physical process, accurate estimation of physical parameters, performance of large-scale parallel computing, and rapid post-processing and visualization have become the bottleneck of its development and application. The rapid development of high performance computing and neural network depth learning technologies provide an opportunities for the breakthrough of ocean models. Ocean surface waves and circulation are the most important physical phenomena in the ocean. Based on the ShenWei E-supercomputer and mass oceanic data, this project will focus on ocean surface wave model and ocean circulation model to increase the model simulating speed including large-scale parallel computing performance and I/O efficient, improve the model simulation by estimating the physical parameter more accurately through the deep learning method, and improve the model post-processing and mass data visualization ability. Through the research of this project, we expect to enhance Chinese ocean model development and application level, promote the integration among physical oceanography, high performance computing and artificial intelligence deep learning, and come into being the large-scale demonstration applications for the ShenWei E-supercomputer.