基于遙感監(jiān)測的黑泥灣筏式海帶養(yǎng)殖區(qū)時空變化研究

打開文本圖片集

（1.Schoolof InformationEngineering，ZhejangOcean UniversityZhoushan316022，China;2.State KeyLaboratoryfSatelite Ocean Environment Dynamics，Second Institute of Oceanography，Ministryof Natural Resources，Hangzhou 310ol2，China;

3.Institute of Marine Instrumentation，Qilu Universityof Technology（Shandong Academy ofSciences），Qingdao 26061， China;4.Shandong Provincial Academician Workstation，Shandong Academy of Sciences， Jinan 250o14，China;

5.Laoshan Laboratory，Qingdao 266237，China）

Abstract：Accurate assessment of the scaleanddistribution of offshore marineaquaculture iscritical for effective management，spatial planning，andecological protection.Thisstudyemployed high-resolutionSentinel-2A/Batellite imagery，a U-Netdeeplearningmodelforautomaticfeatureextraction，andhuman-computerinteractivecorrctiontomap the spatialextent of raft-based kelpfarming in Heiniwan Bay from2016 to2024.Theanalysisrevealedathree-phase development trajectoryintheaquaculture areaover thenine-yearperiod.Spatialdistribution exhibitedastable“north-south agglomerationwith central sparsity”patern.The observed spatiotemporal dynamics reflectthe combined influence of technolgical advancements，policy interventions，andnatural environmental conditions.These findings oferarobust scientific basis foroptimizingaquaculture zoning，adaptive management strategies，and ecological governance in coastal regions with comparable aquaculture practices and environmental settings.

Key Words ： Sentinel-2A/2B;remote sensing;raft-based kelpaquaculture；deeplearning; Heiniwan Bay;spatiotempora evolution

海水養(yǎng)殖是全球水產(chǎn)行業(yè)的重要組成部分，其作為現(xiàn)代農(nóng)業(yè)與漁業(yè)的結(jié)合模式，既能夠滿足人們?nèi)找嬖鲩L的海產(chǎn)品需求，又能有效促進(jìn)沿海地區(qū)的發(fā)展[1]。（剩余11038字）