基于卫星遥感和塔基视频的河流流量监测研究OA北大核心CSTPCD

[Objective]In order to obtain long-term,high-frequency of typical section river flow in the Xijiang River Basin,effec-tively supporting the needs of flood warning and water resources management.[Methods]Unmanned aerial vehicle remote sensing and intelligent unmanned boats are integrated to acquire elevation data above and below the water surface of cross sections,con-structing the high-precision three-dimensional digital river channel model.Furthermore,the width of the river width are extracted based on optical and radar satellite images,and an innovative algorithm is developed to identify the river water level and surface flow velocity by combining tower videos and deep learning models,thereby achieving rapid monitoring of river flow.[Results]The result indicate that the water surface width,obtained through integrating tower video and the three-dimensional digital river channel model.shows good consistency with the monitoring result from satellite remote sensing images,with errors ranging be-tween 1～2 pixels(10～20 m).Due to the influence of climate and riparian vegetation,the error will increase with the increasing of the water surface width to a certain extent..The surface flow velocity of the river is closely related to the underwater topogra-phy.When the river is deep,the surface flow velocity will also undergo significant changes.Compared to flow data obtained through hydrological tests using Acoustic Doppler Current Profilers,the video and remote sensing technology-based flow result are generally consistent,with errors controlled within 2%.[Conclusion]The innovative technology is developed to monitor the flow of typical sections by integration of satellite remote sensing and tower video method and conduct accuracy assessments,providing data support and technical assurance for flow monitoring and flood warning in different scenarios in the Xijiang River Basin.