样本特性对海洋遥感产品真实性检验的定量化影响研究

The purpose of Validation is not only in checking the algorithm of extraction method, but also in providing the accuracy of the remote sensing products to users. The characteristics of the sample, which including sample size, average value, standard deviation, range, the spatial and temporal window size of collocating have an impact on the validation results. In validation activities for ocean remote sensing products, the single characteristic and the small size of the sample can reduce the reliability of the validation results.. In this proposal, we will get the quantitative relationship between the distribution characteristics of sample and the validation results, and the quantitative relationship between the spatial temporal window sizes of collocating and validation by using simulation of theoretical error model for retrieval method of ocean parameter. We will validate the remote sensing products of sea surface temperature and significant wave height against in-situ measurements, and get the quantitative relationships again. In the end, we will compare the results acquired from theoretical model with that acquired from remote sensing products.. The work in this proposal will lay a theoretical and technological foundation of validation of ocean remote sensing products extracted form sensors aboard on Chinese HY-1 and HY-2 series ocean satellites.

真实性检验不仅可对反演算法进行有效性检验，同时还为用户提供产品精度的信息。海洋遥感产品真实性检验的数据样本量大小、均值、标准方差、分布范围等特性以及时空匹配窗口的选择均可对检验结果产生影响。在海洋遥感产品真实性检验活动中，样本特性单一和样本量过小会降低检验结果的可信度。. 本项目使用海洋要素反演的误差理论模型模拟分析结果，确定海洋遥感产品真实性检验最小样本量、样本分布特性对海洋遥感产品真实性检验结果影响的定量关系。同时本项目拟利用海表温度、有效波高等海洋遥感产品为实例，通过选取不同研究区域内具有特定样本特性的数据开展真实性检验，将实例检验结果对理论误差分析模型进行验证。. 该项目研究成果预期可为我国海洋一号（HY-1）、海洋二号（HY-2）系列卫星海洋遥感产品精度评价提供理论和技术支持。

真实性检验不仅可对反演算法进行有效性检验，同时还为用户提供产品精度的信息。本项目利用海表温度、有效波高、海面风场等海洋遥感产品为实例，通过选取不同研究区域内具有特定样本特性的数据开展了其真实性检验，且分析了现场观测数据与卫星海洋遥感产品的时空匹配方法。研究结果表明样本量、样本特性（即均值、方差、最小和最大值）均为影响遥感产品真实性检验结果的因素。本项目获得的真实性检验方法成果预期可为我国海洋一号（HY-1）、海洋二号（HY-2）系列卫星海洋遥感产品精度评价提供支持。