数据与计算发展前沿 ›› 2026, Vol. 8 ›› Issue (2): 3-14.

CSTR: 32002.14.jfdc.CN10-1649/TP.2026.02.001

doi: 10.11871/jfdc.issn.2096-742X.2026.02.001

• 专刊：冰冻圈大数据挖掘分析关键技术及应用 • 上一篇    下一篇

冰冻圈“大数据+AI+模型”耦合研究范式探索

张耀南1,2,*(),刘景琦1,2,3,康建芳1,2,3,南卓铜4,田文彪4,敏玉芳1,2,赵书萍4,王保得5,6   

  1. 1 中国科学院西北生态环境资源研究院， 甘肃 兰州 730030
    2 国家冰川冻土沙漠科学数据中心， 甘肃 兰州 730030
    3 中国科学院大学， 北京 100049
    4 南京师范大学， 江苏 南京 210023
    5 中国科学院新疆生态与地理研究所， 新疆 乌鲁木齐 830011
    6 干旱区生态安全与可持续发展全国重点实验室， 新疆 乌鲁木齐 830011
  • 收稿日期:2025-10-10 出版日期:2026-04-20 发布日期:2026-04-23
  • 通讯作者: *张耀南（E-mail: yaonan@lzb.ac.cn）
  • 作者简介:张耀南，中国科学院西北生态环境资源研究院，研究员，博士生导师，国家冰川冻土沙漠科学数据中心主任，主要研究方向为地学数据工程及数据工程防灾减灾、基于高性能计算环境的地学模型模拟、遥感图像处理及多源数据融合。
    本文主要承担工作为论文构思、论文撰写。
    ZHANG Yaonan, Ph.D., is a professor in Northwest Institute of Eco-Environment and Resources and director of the National Cryosphere Desert Science Data Center. His main research interests include data engineering, disaster prevention and reduction with data engineering, integrated modeling, remote sensing image processing, and multi-source heterogeneous data fusion.
    In this paper, he is mainly responsible for the conception and writing of the paper.
    E-mail: yaonan@lzb.ac.cn
  • 基金资助:
    国家重点研发计划(2022YFF0711704);2022年度新疆交通运输行业科技项目(2022-ZD-006);新疆交投2021年揭榜挂帅科技项目(ZKXFWCG2022060004);新疆交通设计院公司科研基金(KY2022041101);新疆维吾尔自治区科技支疆项目(2024E02030)

Exploration of An Integrated “Big Data+AI+Modeling” Research Paradigm for Cryosphere Studies

ZHANG Yaonan1,2,*(),LIU Jingqi1,2,3,KANG Jianfang1,2,3,NAN Zhuotong4,TIAN Wenbiao4,MIN Yufang1,2,ZHAO Shuping4,Wang Baode5,6   

  1. 1 Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences, Lanzhou, Gansu 730030, China
    2 National Cryosphere Desert Data Center, Lanzhou, Gansu 730030, China
    3 University of Chinese Academy of Sciences, Beijing 100049, China
    4 Nanjing Normal University, Nanjing, Jiangsu 210023, China
    5 Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, Chinese Academy of Sciences, Urumqi, Xinjiang 830011, China
    6 State Key Laboratory of Ecological Security and Sustainable Development in Arid Areas, Urumqi, Xinjiang 830011, China
  • Received:2025-10-10 Online:2026-04-20 Published:2026-04-23

RichHTML

6

PDF

24

摘要：

【目的】 冰冻圈作为地球圈层系统的重要部分，其多要素非线性联动、跨圈层耦合及灾害链对全球气候、水文循环及生态安全具有深刻影响。传统物理模型在多尺度、多要素联动表征冰冻圈变化方面存在局限，亟需建立新的科学研究范式来满足冰冻圈大尺度变化监测、快速演变解析及灾害预警等实际需求。【方法】 本文提出了“大数据+AI+模型”冰冻圈研究新范式，研发了支持该范式研究的全球冰冻圈研究引擎（Global Cryosphere research Engine，GCE），形成了数据管理、模型构建、AI算法应用、任务流程编排等全链条一体化平台，具备多源异构数据融合、跨语言模型协同、高性能计算调度与智能分析能力。【结论】 基于GCE平台，实现了“物理模型嵌入AI”“AI嵌入物理模型”“数据驱动耦合AI的参数优化”模式的构建，并围绕青藏高原土壤水分模拟试验，建立了“观测数据+深度学习（分层LSTM）+物理模型（Noah-MP）”多要素联动框架，土壤水分分层模拟结果与实测值的平均相关系数超过0.8，显著高于Noah模型。研究表明，物理模型融合人工智能的研究范式可有效刻画冰冻圈过程，验证了GCE平台的可用性，可为生态环境、冰冻圈灾害等新范式研究提供环境支持，推动冰冻圈向跨区域多要素联动的智能化研究深入。

关键词: 冰冻圈, 大数据, 人工智能, 模型, 冰冻圈研究引擎GCE

Abstract:

[Objective] As a vital component of Earth’s system, the cryosphere profoundly impacts global climate, hydrological cycles, and ecological security through multi-element nonlinear interactions, cross-sphere couplings, and long disaster chains. Traditional physical models exhibit limitations in characterizing multi-scale cryospheric changes, failing to meet practical needs for large-scale monitoring, rapid evolution analysis, and hazard early-warning. [Methods] This paper proposes a “Big Data+AI+Model” research paradigm for cryospheric studies and develops the Global Cryosphere research Engine (GCE) to support this framework. The GCE forms an integrated platform encompassing data management, model construction, AI algorithm application, and task workflow orchestration, with capabilities for multi-source heterogeneous data fusion, cross-language model collaboration, high-performance computing scheduling, and intelligent analysis. [Conclusions] Based on the GCE platform, model construction approaches such as “embedding AI into physical models”, “embedding physical models into AI”, and “data-driven parameter optimization coupled with AI” have been implemented. Focusing on soil moisture simulation experiments over the Tibetan Plateau, a novel multi-element interaction framework integrating “observation data+deep learning (hierarchical LSTM)+physical model (Noah-MP)” was developed. The average correlation between simulated and observed layered soil moisture exceeded 0.8, significantly outperforming the Noah model. The results demonstrate the feasibility of a research paradigm that combines data-driven approaches, physical model support, and artificial intelligence to effectively characterize cryospheric processes, while also validating the utility of the GCE platform. This framework can provide a supportive environment for new research paradigms in ecological environments and cryospheric disasters, promoting intelligent cryospheric research with cross-regional multi-element interactions.

Key words: cryosphere, big data, AI, model, GCE (Global Cryosphere research Engine)