源区黑潮流量变化对局地气候、海洋环流及生态系统有重要影响，开展其可预报性和目标观测研究具有重要意义。初始误差是导致源区黑潮流量预测结果不确定性的重要原因，海洋背景状态在误差发展过程中起着关键作用。因而在不同季节海洋背景状态下探究初始误差对源区黑潮流量变化预报的影响有重要意义。为此，本项目拟基于复杂海洋模式（ROMS），利用条件非线性最优扰动（CNOP）方法开展如下研究：（1）在不同季节，利用CNOP方法探究初始误差对源区黑潮流量变化预测的影响，探讨误差增长机制及海洋背景状态季节差异对误差发展的影响；（2）基于CNOP型误差的空间结构特征识别源区黑潮流量变化的目标观测敏感区，并从应用角度出发，于敏感区内设计适应性观测网。本项目的实施既有助于弄清初始误差导致源区黑潮流量季节性变化预测不确定性的原因和机制，又能为该海区内观测系统的优化设计提供方法指导。

As the seasonal variation of the upstream Kuroshio transport has important impacts on regional climate, ocean circulation and ecosystem, it is of great significance to study its predictability and targeted observation problem. Recent studies show that initial errors play a vital role in affecting the upstream Kuroshio transport forecast and ocean background state affects the evolution of initial errors significantly. Therefore, it is worth investigating how initial errors affect the upstream Kuroshio transport prediction under different ocean background states. By using the Regional Ocean Modeling System (ROMS) and the conditional nonlinear optimal perturbation (CNOP) method, this project will investigate the following problems. Firstly, how initial errors in each season affect the prediction of the upstream Kuroshio transport will be investigated. The error-growth mechanism and the impacts of ocean background states on the evolution of initial errors will be specifically discussed, respectively. Then, the sensitive areas of targeted observation on improving the upstream Kuroshio transport prediction will be determined through the spatial structures of the CNOP-type errors. Adaptive observation networks will be further designed within the identified sensitive areas. This study is helpful to clarify the reason and mechanism causing the prediction uncertainties of the seasonal variation of the upstream Kuroshio transport and will provide guidance on operationally implementing adaptive observation in this region.