文献类型：期刊文献

英文题名：The fastest growing initial error and identification of sensitive area for targeted observation in predicting the Kuroshio intrusion into the South China Sea with a high-resolution regional ocean model

作者：Liang, Peng[1,2,3];Liang, Yonghao[1];Wang, Qiang[4,5];Yang, Lina[1,2,3];Zhang, Tianyu[1,2,3]

机构：[1]Guang dong Ocean Univ, Coll Ocean & Meteorol, Lab Coastal Ocean Variat & Disaster Predict, Zhanjiang, Peoples R China;[2]Resources & Environm Continental Shelf Sea & Deep, Key Lab Climate, Zhanjiang, Peoples R China;[3]Minist Nat Resources, Key Lab Space Ocean Remote Sensing & Applicat, Beijing, Peoples R China;[4]Hohai Univ, Key Lab Marine Hazards Forecasting, Minist Nat Resources, Nanjing, Peoples R China;[5]Hohai Univ, Coll Oceanog, Nanjing, Peoples R China

年份：2026

卷号：201

外文期刊名：OCEAN MODELLING

收录：SCI-EXPANDED(收录号：WOS:001689998600001)、、EI(收录号：20260620036042)、Scopus(收录号：2-s2.0-105029414176)、WOS

基金：This study is supported by the National Natural Science Foundation of China (Grants 42288101, 42006023, 42476219, 42476192 and 41706033), Innovative Programs for Regular Higher Education Institutions in Guangdong Province (2024KTSCX200), National Foreign Experts Program grant S20240134, the program for scientific research start-up funds of Guangdong Ocean University (Grants R20020, R20023), the team project funding of scientific research innovation for universities in Guangdong Province (2023KCXTD015), Guangdong Provincial Observation and Research Station for Tropical Ocean Environment in Western Coastal Waters (GSTOEW), and the Modern Education Technology Center of Guangdong Ocean University.

语种：英文

外文关键词：Kuroshio intrusion into the South China Sea; Initial errors; Error growth dynamics; Sensitive area

外文摘要：Kuroshio intrusion (KI) is a critical linkage between the Pacific and the South China Sea (SCS), profoundly influencing the variability of marine dynamical and ecological processes of the SCS. Due to the complex mechanism and the lack of predictability study on KI, the accuracy of KI prediction remains limited. This study obtains the fastest growing initial errors (FGIEs) of KI using the Regional Ocean Modelling System (ROMS) and conditional nonlinear optimal perturbation (CNOP) method. Specifically, the CNOP, which is an effective method in calculating FGIEs in a nonlinear system, refers to the perturbation that can lead to the maximum of an objective function at a target time under certain constraints. The calculation results reveal two types of FGIEs with similar spatial patterns but opposite signs. When superimposed on the background field, both types of errors exhibit rapid growth and northwestward propagation. At prediction time, the CNOP+ (with positive sea surface height error) and CNOP- (with negative sea surface height error) errors respectively cause significant overestimation and underestimation of KI. Notably, CNOP- errors may even lead to complete failure in predicting the occurrence of KI. The rapid error growth primarily originates from barotropic instability induced by the zonal velocity shear of the reference state. Sensitive areas for targeted observations, identified through vertical integration of initial total energy error, extend northwestward from the southern Luzon Strait to the interior SCS, centered near 120.5 degrees E, 20 degrees N. Remarkably, removing initial errors within this sensitive area (covering merely 0.1 % of the total model domain) can improve KI prediction accuracy most effectively, by 25 %similar to 38 %. This research provides an effective guidance for the design of targeted observation strategies, having great significance in improving the prediction skill of KI.