大河影响下的边缘海水体二氧化碳分压卫星遥感反演研究

River-dominated Ocean Margins (RioMar) are one of the key regions to influence the uncertainty of the global carbon budget estimation. Satellite remote sensing is an effective method to monitor these highly dynamic systems. Currently, the remote sensing models of partial pressure of aquatic carbon dioxide (pCO2) are generally based on the empirical regression, which are difficult to implement the inversion of pCO2 in the complex marginal seas. In this project, we will take the Changjiang River-dominated East China Sea (ECS) as the target region, and develop the mechanistic-based semi-analytical remote sensing algorithm (MSAA) of the pCO2 based on the calculations of marine carbonate system, ocean optical properties and satellite remote sensing. Firstly, we will analyze the seasonality of major controlling mechanisms of the pCO2 in the ECS, then accomplish the semi-analytical parameterization for individual pCO2 components contributed by each process or controlling factor using satellite-derived variables, and develop the MSAA algorithm of pCO2. Then, we will generate monthly pCO2 and air-sea carbon dioxide flux in the ECS using satellite data in 20-year time series, and analyze their spatial-temporal variations and responses to the climate changes. Finally, we will try to apply the MSAA algorithm to the two marginal seas influenced by the Mississippi River and Amazon River, to exam the capability of the MSAA algorithm to global river-dominated ocean margin. The implement of this project can not only promote the cross-discipline study between the satellite remote sensing and marine chemistry, but also deepen the insight of spatial-temporal evolution of the air-sea dioxide carbon flux in the complex marginal seas and eliminate the uncertainty of carbon budget, especially the understanding of their responses to the climate changes.

大河影响下的边缘海是减少全球碳收支估算不确定性的关键区域之一，需要高时空分辨率及长期稳定的遥感监测。目前基于统计回归建立的海水CO2分压（pCO2）遥感模型通常难以描述复杂边缘海pCO2的时空变化。本项目以长江冲淡水影响下的东海陆架为主要研究区域，结合海水碳酸盐系统和水体光学特性，研究基于控制机制分析的海水pCO2遥感反演模型。通过分析东海不同季节及水团pCO2的主要控制机制，建立各主控因子引起pCO2变化的遥感量化模型，构建东海水体pCO2遥感半解析算法。在此基础上，制作东海陆架20年序列的月平均海水pCO2及海-气CO2通量遥感产品集，分析其时空变化及对气候变化的响应。同时，将该半解析算法应用到亚马逊河及密西西比河影响的边缘海系统，以检验算法的推广能力。项目可促进卫星遥感与海洋碳化学的多学科交叉研究，还能深化对复杂边缘海系统碳收支时空演变，特别是对气候变化响应的认识。

大河影响下边缘海海-气CO2通量具有高动态变化的特点，是减少全球碳收支评估不确定性的关键区域之一，需要高时空分辨率及长期稳定的遥感监测。相对于大洋水体，边缘海系统水体CO2分压受众多因子调控，其遥感反演具有较大的难度。本项目围绕长江冲淡水影响下的东海陆架水体CO2分压遥感反演问题，开展了系统、深入的研究。创新提出了基于控制因子量化的海水CO2分压遥感半分析算法模型（MeSAA），不仅考虑了陆源影响，且通过主控因子影响的叠加实现同一模型在全海域的应用，克服了传统多模型切换导致的产品斑块问题。MeSAA算法模型的半分析特性，可拓展到不同边缘海系统，具有广阔的应用前景。MeSAA算法模型为边缘海海-气CO2通量遥感反演提供了崭新的框架体系，已成功在珠江冲淡水系统、密西西比河冲淡水系统，以及在大洋主控型边缘海白令海成功应用。基于MeSAA算法模型，制作了近20年来东海pCO2遥感产品，分析了其时空变化规律，且已在SatCO2平台上(www.SatCO2.com)发布共享。项目研究在海洋学和遥感学主流期刊《JGR-Ocean》、《Remote Sensing》等发表论文24篇，其中SCI论文19篇，MeSAA算法文章被国际海洋水色协调组织（IOCCG）门户网站作亮点报导。共培养研究生9人，项目负责人2017年获浙江省自然科学基金杰出青年项目资助，2018年入选首批浙江省 “万人计划”领军人才称号。