项目名称：山洪灾害信号早期识别与准确预警技术装备

English Name: Early identification and accurate early warning technology and equipment for mountain torrent disasters

 执行年限/Time: 2022.11-2025.10

项目简介：

本项目研发目标是:突破不同特征流域暴雨山洪形成机理，实现基于自主可控技术的山洪关键产汇流参数和过程的空天地一体化快速采集与实时传输，提升致洪暴雨精细化监测和短临预报技术水平及山洪模型性能，实现山洪多尺度滚动预报和精准动态分级预警，形成可复制可推广的技术体系，为提升我国山洪预报预警水平提供科技支撑。

针对山洪精准动态预警中的关键难点，重点回答3个关键科学问题:不同特征流域暴雨山洪产流机制，山洪升尺度规律和本构关系，不同尺度和类型致洪暴雨天气系统关键物理因子和变化规律;解决7项关键技术问题:山区复杂环境产汇流参数空天地一体化精细快速测量技术，土壤水分和坡面产流空天地一体化精细监测和实时传输技术，地表水-土壤水-地下水耦合的多过程多尺度高效高分辨率山洪模拟技术，致洪暴雨超高时空分辨率雷达反演技术，物理机制和深度学习相结合的致洪暴雨精准短临预报技术，暴雨山洪精准动态分级预警指标确定技术，泥石流装配式拦排技术。

围绕以上关键问题，精选不同气候区山洪实验流域开展强化观测实验，揭示山洪机理并作为研发技术的整体验证流域。立足自主可控研发山洪关键参数和过程的空天地一体化监测技术、高效高分辨率水文水动力耦合模型、致洪暴雨超高分辨率监测和多尺度预报技术，并开展示范应用。具体包括5项研究内容:山洪关键产汇流参数和过程的空天地一体化监测技术装备，不同气候区山洪形成机理与本构关系，高效高分辨率山洪模型及其智能建模技术，不同气候区致洪暴雨精细监测和多尺度预报技术，山洪多尺度精准化预报预警系统和应用示范。

本项目将融合相关研究的最新进展，以空天地一体化监测、强化观测实验、机器学习等手段为支撑，按照“数据底板-机理解析-精准预报-动态预警”的总体思路进行研发。数据底板方面，将研发山洪关键产汇流参数和过程的空天地一体化精细测量技术装备，对包括高精度地表高程、植被冠层、糙率、土壤水分等在内的山洪关键参数和变量进行精细化监测;机理解析方面，将通过在不同实验区的观测研究，揭示不同流域山洪特征、发生规律及主导产流机制，揭示不同尺度和类型致洪暴雨天气系统的关键物理因子和变化规律;精准预报方面，将基于精细化监测和机理研究成果，研发高分辨率致洪暴雨预报技术和高效高分辨率山洪模拟技术，实现对山洪沟道不同断面水深和流速的准确快速计算:动态预警方面，以小流域为单元对可能山洪过程进行海量计算，以降雨量为指标确定动态分级预警阈值，融合精细化土壤水分监测开展实时动态预警。

本项目将创新不同特征流域山洪形成机制、致洪暴雨降水系统变化规律等重要科学问题的认知，研发山洪灾害多尺度滚动预报和动态分级预警系统，编写山洪灾害准确监测和早期预警技术指南，形成可复制可推广的技术体系，为提升我国山洪灾害防御水平提供坚实的科技支撑，具有显著的科学价值和巨大的经济社会效益。

Project Introduction:

The research and development objective of this project is to break through the formation mechanism of torrential mountain floods in different characteristic basins, achieve rapid collection and real-time transmission of key production and convergence parameters and processes of mountain floods based on independently controllable technology through an integrated air, space and ground system, enhance the technical level of fine monitoring and short-term and near-term forecasting of flood-causing rainstorms and the performance of mountain flood models, realize multi-scale rolling forecasting and precise dynamic classification early warning of mountain floods, form a replicable and scalable technical system, and provide scientific and technological support for improving the level of mountain flood forecasting and early warning in China.

To address the key difficulties in the precise and dynamic early warning of mountain floods, three critical scientific questions are focused on: the runoff generation mechanism of mountain floods caused by heavy rain in different characteristic basins, the upscaling laws and constitutive relations of mountain floods, and the key physical factors and variation laws of different scales and types of rainstorm weather systems that cause floods. Seven key technical issues are also to be resolved: the fine and rapid measurement technology of runoff generation and convergence parameters in mountainous complex environments through an integrated space-air-ground approach, the fine monitoring and real-time transmission technology of soil moisture and slope runoff through an integrated space-air-ground approach, the high-efficiency and high-resolution multi-process and multi-scale simulation technology of mountain floods with the coupling of surface water, soil water and groundwater, the ultra-high spatio-temporal resolution radar inversion technology of rainstorms causing floods, the precise short-term and nowcasting technology of rainstorms causing floods by combining physical mechanisms and deep learning, the determination technology of precise dynamic grading warning indicators for rainstorm floods, and the prefabricated interception and diversion technology for debris flows.

Centering around the above key issues, selected flash flood experimental basins in different climate zones will be carried out for intensive observation experiments to reveal the mechanism of flash floods and serve as the overall verification basins for the research and development of technologies. Based on self-control, we will develop the space-air-ground integrated monitoring technology for key flash flood parameters and processes, the high-efficiency and high-resolution hydrological and hydrodynamic coupling model, the ultra-high-resolution monitoring and multi-scale forecasting technology for flood-causing rainstorms, and conduct demonstration applications. Specifically, it includes five research contents: the space-air-ground integrated monitoring technology and equipment for key flash flood runoff generation and concentration parameters and processes, the formation mechanism and constitutive relations of flash floods in different climate zones, the high-efficiency and high-resolution flash flood model and its intelligent modeling technology, the fine monitoring and multi-scale forecasting technology for flood-causing rainstorms in different climate zones, and the multi-scale precise flash flood forecasting and early warning system and application demonstration.

This project will integrate the latest advancements in related research, and be supported by means such as space-air-ground integrated monitoring, intensive observation experiments, and machine learning. It will be developed following the general approach of "data foundation - mechanism analysis - precise forecasting - dynamic early warning".Data Foundation: We will develop space-air-ground integrated fine measurement technology and equipment for key flash flood runoff generation and concentration parameters and processes. This will enable the fine - grained monitoring of key flash flood parameters and variables, including high - precision surface elevation, vegetation canopy, roughness, soil moisture, etc.Mechanism Analysis: Through observational research in different experimental areas, we will reveal the characteristics, occurrence patterns, and dominant runoff - generation mechanisms of flash floods in different basins. Additionally, we will uncover the key physical factors and variation patterns of flood - causing rainstorm weather systems at different scales and of different types.Precise Forecasting: Based on the results of fine - grained monitoring and mechanism research, we will develop high - resolution flood - causing rainstorm forecasting technology and high - efficiency, high - resolution flash flood simulation technology. This will enable the accurate and rapid calculation of water depth and flow velocity at different cross - sections of flash - flood gullies.Dynamic Early Warning: Taking small basins as units, we will conduct massive calculations on possible flash - flood processes. We will determine dynamic graded early - warning thresholds using rainfall as an indicator, and integrate fine - grained soil moisture monitoring to carry out real - time dynamic early warning.

This project will innovate the understanding of important scientific issues such as the formation mechanism of flash floods in basins with different characteristics and the variation laws of flood-causing rainstorm precipitation systems. It will also develop a multi-scale rolling forecasting and dynamic graded early warning system for flash flood disasters, compile technical guidelines for the accurate monitoring and early warning of flash flood disasters, form a replicable and promotable technical system, provide solid scientific and technological support for improving the defense level against flash flood disasters in China, and have significant scientific value as well as huge economic and social benefits.