The international Argo program,a global observational array of nearly 4000 autonomous profiling floats initiated in the late 1990s,which measures the water temperature and salinity of the upper 2000 m of the global oc...The international Argo program,a global observational array of nearly 4000 autonomous profiling floats initiated in the late 1990s,which measures the water temperature and salinity of the upper 2000 m of the global ocean,has revolutionized oceanography.It has been recognized one of the most successful ocean observation systems in the world.Today,the proposed decade action“OneArgo”for building an integrated global,full-depth,and multidisciplinary ocean observing array for beyond 2020 has been endorsed.In the past two decades since 2002,with more than 500 Argo deployments and 80 operational floats currently,China has become an important partner of the Argo program.Two DACs have been established to process the data reported from all Chinese floats and deliver these data to the GDACs in real time,adhering to the unified quality control procedures proposed by the Argo Data Management Team.Several Argo products have been developed and released,allowing accurate estimations of global ocean warming,sea level change and the hydrological cycle,at interannual to decadal scales.In addition,Deep and BGC-Argo floats have been deployed,and time series observations from these floats have proven to be extremely useful,particularly in the analysis of synoptic-scale to decadal-scale dynamics.The future aim of China Argo is to build and maintain a regional Argo fleet comprising approximately 400 floats in the northwestern Pacific,South China Sea,and Indian Ocean,accounting for 9%of the global fleet,in addition to maintaining 300 Deep Argo floats in the global ocean(25%of the global Deep Argo fleet).A regional BGC-Argo array in the western Pacific also needs to be established and maintained.展开更多
海洋温度数据在全球海洋观测和气候研究中发挥着关键作用,质量控制对于确保这些数据的可靠性十分关键,然而,目前在大数据集上的异常数据召回率尚不理想。文章基于Argo温度数据,提出一种基于规则集和多层感知机(rule set and multilayer ...海洋温度数据在全球海洋观测和气候研究中发挥着关键作用,质量控制对于确保这些数据的可靠性十分关键,然而,目前在大数据集上的异常数据召回率尚不理想。文章基于Argo温度数据,提出一种基于规则集和多层感知机(rule set and multilayer perceptron,RS-MLP)的质量控制方法。首先对13种机器学习模型进行对比分析,从中筛选出最优机器学习模型,然后设计了由6种基于规则的质量控制检查模块组成的规则集,最后集成规则集和最优机器学习模型构建出RS-MLP方法,并以南海区域的Argo数据为例评估方法性能。研究结果表明:RS-MLP在351746条温度数据的测试集中真阳性率(true positive rate,TPR)、真阴性率(true negative rate,TNR)和接受者操作特性(receiver operating characteristic,ROC)曲线下面积(area under the curve,AUC)依次能达到93%、96%和95%,并在不同深度层次上的异常数据召回率比较稳定,具有优秀的质量控制性能。展开更多
基于不同机构(SIO、JAM和EN4)发布的多套Argo数据集,分析盐度漂移的时空变化特征。结果表明,2016年以后,3家机构发布的盐度数据集都存在明显的系统性漂移,且不同深度层的盐度漂移幅度差异较大。为此,基于赤池信息量准则(Akaike informat...基于不同机构(SIO、JAM和EN4)发布的多套Argo数据集,分析盐度漂移的时空变化特征。结果表明,2016年以后,3家机构发布的盐度数据集都存在明显的系统性漂移,且不同深度层的盐度漂移幅度差异较大。为此,基于赤池信息量准则(Akaike information criterion,AIC)提出面向特定深度层盐度漂移的多项式修正方法。该方法能有效修正0~2000m深度范围内Argo数据集的盐度偏差,使得2005~2021年的GMSL预算平衡偏差减少约43%。盐度漂移具有复杂的空间关联性,全球盐容海平面变化趋势存在显著的空间分布差异,北大西洋区域最为显著,修正后其空间分布趋于一致。展开更多
本文针对海洋数据大幅增长,海量数据信息管理存在困难,数据质量存在参差不齐的情况,提出Argo数据质量控制和数据集制作的一整套方案,显著提升了海洋信息管理行业工作效率。点评人:刘嘉玥,天津大学海洋科学与技术学院副教授,硕士研究生导...本文针对海洋数据大幅增长,海量数据信息管理存在困难,数据质量存在参差不齐的情况,提出Argo数据质量控制和数据集制作的一整套方案,显著提升了海洋信息管理行业工作效率。点评人:刘嘉玥,天津大学海洋科学与技术学院副教授,硕士研究生导师,研究领域:海洋战略、海洋信息。目前海洋数据采集范围最广,数据量增长最快的国际Argo(ARRAY for REAL-TIME GEOSTROPHIC OCEANOGRAPHY)计划是获取海洋数据的重要途径。该计划通过在全球海洋中布放Argo剖面浮标,使其组成一个实时和高分辨率的海洋观测网,并借助卫星定位和通信系统,实时、准确、大范围地获取全球海洋上层(0~2000m)的海水温度、盐度剖面资料。面对如此巨大体量的海洋数据,如何做好数据的信息化管理至关重要。展开更多
基金The National Natural Science Foundation of China under contract Nos 42122046,42076202,U1811464 and 4210060098the Project Supported by Laoshan Laboratory under contract No.LSKJ202201500the Project Supported by Southern Marine Science and Engineering Guangdong Laboratory(Zhuhai)under contract No.SML2021SP102.
文摘The international Argo program,a global observational array of nearly 4000 autonomous profiling floats initiated in the late 1990s,which measures the water temperature and salinity of the upper 2000 m of the global ocean,has revolutionized oceanography.It has been recognized one of the most successful ocean observation systems in the world.Today,the proposed decade action“OneArgo”for building an integrated global,full-depth,and multidisciplinary ocean observing array for beyond 2020 has been endorsed.In the past two decades since 2002,with more than 500 Argo deployments and 80 operational floats currently,China has become an important partner of the Argo program.Two DACs have been established to process the data reported from all Chinese floats and deliver these data to the GDACs in real time,adhering to the unified quality control procedures proposed by the Argo Data Management Team.Several Argo products have been developed and released,allowing accurate estimations of global ocean warming,sea level change and the hydrological cycle,at interannual to decadal scales.In addition,Deep and BGC-Argo floats have been deployed,and time series observations from these floats have proven to be extremely useful,particularly in the analysis of synoptic-scale to decadal-scale dynamics.The future aim of China Argo is to build and maintain a regional Argo fleet comprising approximately 400 floats in the northwestern Pacific,South China Sea,and Indian Ocean,accounting for 9%of the global fleet,in addition to maintaining 300 Deep Argo floats in the global ocean(25%of the global Deep Argo fleet).A regional BGC-Argo array in the western Pacific also needs to be established and maintained.
文摘海洋温度数据在全球海洋观测和气候研究中发挥着关键作用,质量控制对于确保这些数据的可靠性十分关键,然而,目前在大数据集上的异常数据召回率尚不理想。文章基于Argo温度数据,提出一种基于规则集和多层感知机(rule set and multilayer perceptron,RS-MLP)的质量控制方法。首先对13种机器学习模型进行对比分析,从中筛选出最优机器学习模型,然后设计了由6种基于规则的质量控制检查模块组成的规则集,最后集成规则集和最优机器学习模型构建出RS-MLP方法,并以南海区域的Argo数据为例评估方法性能。研究结果表明:RS-MLP在351746条温度数据的测试集中真阳性率(true positive rate,TPR)、真阴性率(true negative rate,TNR)和接受者操作特性(receiver operating characteristic,ROC)曲线下面积(area under the curve,AUC)依次能达到93%、96%和95%,并在不同深度层次上的异常数据召回率比较稳定,具有优秀的质量控制性能。
文摘基于不同机构(SIO、JAM和EN4)发布的多套Argo数据集,分析盐度漂移的时空变化特征。结果表明,2016年以后,3家机构发布的盐度数据集都存在明显的系统性漂移,且不同深度层的盐度漂移幅度差异较大。为此,基于赤池信息量准则(Akaike information criterion,AIC)提出面向特定深度层盐度漂移的多项式修正方法。该方法能有效修正0~2000m深度范围内Argo数据集的盐度偏差,使得2005~2021年的GMSL预算平衡偏差减少约43%。盐度漂移具有复杂的空间关联性,全球盐容海平面变化趋势存在显著的空间分布差异,北大西洋区域最为显著,修正后其空间分布趋于一致。
文摘本文针对海洋数据大幅增长,海量数据信息管理存在困难,数据质量存在参差不齐的情况,提出Argo数据质量控制和数据集制作的一整套方案,显著提升了海洋信息管理行业工作效率。点评人:刘嘉玥,天津大学海洋科学与技术学院副教授,硕士研究生导师,研究领域:海洋战略、海洋信息。目前海洋数据采集范围最广,数据量增长最快的国际Argo(ARRAY for REAL-TIME GEOSTROPHIC OCEANOGRAPHY)计划是获取海洋数据的重要途径。该计划通过在全球海洋中布放Argo剖面浮标,使其组成一个实时和高分辨率的海洋观测网,并借助卫星定位和通信系统,实时、准确、大范围地获取全球海洋上层(0~2000m)的海水温度、盐度剖面资料。面对如此巨大体量的海洋数据,如何做好数据的信息化管理至关重要。
