Sea surface currents observed by high-frequency (HF) radars have been widely used in ocean circulation research. In this study, hourly sea surface currents observed by the Taiwan Coastal Ocean Dynamics Applications Radar (CODAR) system from 2015 to 2019 were analyzed by the empirical orthogonal function (EOF) analysis to reveal the characteristics of the sea surface currents around Taiwan Island. The study area is divided into two regions, the Kuroshio region east of Taiwan Island and the Taiwan Strait west of Taiwan Island. In the Kuroshio region, the first EOF mode shows that the Kuroshio is characterized by higher current speeds with greater variability in summer. The second and third EOF modes present a dipole eddy pair and single eddy impingement on the Kuroshio during different periods. The seasonal variation of the dipole eddy pair indicates that the cyclonic/anticyclonic eddy on the north/south side appears more frequently in summer. Single eddy impingement occurs at multiple periods, including daily, intraseasonal, interseasonal, and annual periods. For the Taiwan Strait, the first EOF mode displays the tide signals. The tides enter the Taiwan Strait from the north and south, forming strong sea surface currents around the northern tip of Taiwan Island and the Penghu Archipelago. The second EOF mode exhibits the seasonal changes of the sea surface currents driven by the monsoon winds. The sea surface currents in the northern Taiwan Strait are relatively strong, possibly due to the narrow and shallow terrain there. The high spatiotemporal resolution of sea surface currents derived from CODAR observations provide more detailed characteristics of sea surface circulation around Taiwan Island.

This study aimed to examine the variations in vorticity in the northern South China Sea (NSCS) resulting from the cut‐off of the Kuroshio main path east of Taiwan Island. Absolute dynamic topography and satellite altimeter eddy tracking data from 1993 to 2021 were employed to analyze the momentum ratio of the Kuroshio and mesoscale eddies concerning Kuroshio cut‐off events through their interaction. The results revealed 12 events where mesoscale cyclonic eddies east of Taiwan Island cut off the Kuroshio main path, resulting in intensified loop currents in the Luzon Strait and increased negative vorticity in the NSCS. Six of these events occurred between 1998 and 2013, coinciding with the negative Pacific Decadal Oscillation (PDO) regime. During this period, negative vorticity triggered by Kuroshio cut‐off events rose by 79% in the NSCS, compared to 34% during the positive PDO regime. The Kuroshio interacted with mesoscale cyclonic eddies east of Taiwan for 28 days during the negative PDO regime and 41 days during the positive PDO regime. The Kuroshio velocity strengthened east of Luzon Island but weakened east of Taiwan Island during the negative PDO regime, elucidating the observed disparity. The average duration between the occurrence of mesoscale cyclonic eddy impinging on the Kuroshio east of Taiwan Island and the peak negative vorticity observed in the NSCS was approximately 28 days. This was comparable to data obtained using field measurements.

The relationship between wind and waves has been extensively studied over time. However, understanding the local wind and wave relationship remains crucial for advancing renewable energy development and optimizing ocean management strategies. This study used wind and wave data collected by the ten weather buoys in the waters surrounding Taiwan to analyze regional sea states. The relationship between wind speed and significant wave height (SWH) was examined using regression analysis. Additionally, machine learning techniques were employed to assess the relative importance of features contributing to SWH growth. The regression analysis revealed that SWH in the waters surrounding Taiwan was not fully developed, with notable discrepancies observed between the waters east and west of Taiwan. According to the power law formula describing the relationship between wind speed and SWH, the eastern waters exhibited a larger prefactor coupled with a smaller scaling exponent, while the western waters manifested a converse parametric configuration. Through an evaluation of four machine learning algorithms, it was determined that wind speed is the most influential factor driving these regional differences, especially in the waters west of Taiwan. Beyond wind speed, air pressure or temperature emerged as the secondary feature factor governing wind–wave interactions in the waters east of Taiwan.

Although the issue of the global warming hiatus during 1998-2013 is still under debate, the surface velocity of the Kuroshio has been shown to decrease east of Taiwan but increase east of Luzon during the global warm hiatus period. In this study, we revealed a significant increase in cut-off events on the Kuroshio main path east of Taiwan during the hiatus period. This is probably related to the interaction between the weakening Kuroshio off the east of Taiwan and the westward propagation of the mesoscale cyclone eddy from the western Pacific Ocean. The cut-off event of the main path of the Kuroshio is determined using the momentum ratio of the Kuroshio to the eddy. If the ratio is less than one, the eddy dominates and the Kuroshio is cut off. Sea surface velocities derived from satellite altimeter data for these events in the Kuroshio region east of Taiwan and the Kuroshio intrusion region west of the Luzon Strait were then analyzed by empirical orthogonal functions. The results showed that the Kuroshio intrusion to the west of the Luzon Strait increased after an average of about one month after the main Kuroshio off east of Taiwan was cut off. This appears to be reverse feedback from the Kuroshio downstream to upstream.

When the Kuroshio passes through the Luzon Strait from its upstream east of Luzon Island to its downstream east of Taiwan, there are three types of possible routes. First of them is the western component, also known as the western branch, which can be further divided into looping path and leaking path (the Kuroshio intruding into the South China Sea). The second type is the mainstream connecting the eastern Luzon Island and the eastern Taiwan Island, which has the pattern nearly as same as the long-term mean of the Kuroshio path in this region. Lastly, the third type of route is the eastern branch that is to be focused by this study. Its definition is that the east components of the current around the Kuroshio's route are greater than the north components and then taking the material away from the Kuroshio main stream in the eastward direction. Therefore, to find the possible routes of the Kuroshio and make the numbers of trajectories of each simulation to be fairly same with the other days, we use OpenDrift (an open-source Python-based framework for Lagrangian particle modeling) as a tool to simulate the trajectories of the Kuroshio started at a given position 18.375°N and 122.875°E from 1993 to 2020. The input data that used as simulation is the geostrophic current derived from altimeter data provided by CMEMS from January 1993 to December 2021. The spatial and temporal resolutions of the input data are 0.25° and one-day, respectively. The results revealed that the accumulated numbers of trajectories as the type of the eastern branch of the Kuroshio would be more frequently during March to June. Meanwhile, the averaged wind stress curl (WSC) of the 10-m wind field data from NCEP/NCAR Reanalysis 1 with a 4-times daily temporal resolution and a 1.875° horizontal resolution was calculated. The difference between the monthly WSC and the annual mean WSC over the entire Pacific Ocean from 1993 to 2020 showed that there is a significant eastward transport along the 20°N-21°N latitude in May. This implies that the eastern branch of the Kuroshio may be caused by WSC.

The Luzon Strait serves as an entrance of the seawater from the Pacific Ocean to the South China Sea (SCS). When the oligotrophic Kuroshio intrusion into the Luzon Strait encounters with the sea surface chlorophyll-a (SChl-a) discharged from the river estuary in the northern Luzon Island, this patch of SChl-a may be transported by the Kuroshio water from the coastal region westward to the northern SCS. During the strong 2015–2016 El Niño event, the Kuroshio intruded into the SCS through the Luzon Strait on 25 and 26 December 2015. Simultaneously, a patch of SChl-a was directed from the northern coast of Luzon Island to the northern SCS along the rotational trajectory of a series of oceanic mesoscale eddies, ultimately resulting in its diffusion. However, measurements from geostrophic currents from satellite altimeter data cannot accurately depict the current fields near the coasts, where the targeted SChl-a patch generated, due to the shallow topography of the Babuyan Channel and complicated Babuyan Islands acting as barriers. The Maximum Cross-Correlation (MCC) method is thereby applied to address this issue by capturing the moving pattern of the target SChl-a patch from the concurrent sea surface temperature (SST) from satellite observations (MODIS-Aqua Level 3) and analysis data (GHRSST Level 4), as well as SChl-a images from satellite observations (MODIS-Aqua Level 3) and merged datasets (OC-CCI and GlobColour). This study investigated the average magnitude of MCC-derived westward current vectors in the Babuyan Channel (18.25°N–18.75°N, 121°E–122°E) after the Kuroshio intrusion occurred in a looping path, Vwest, based on daily GHRSST data (0.01°×0.01°). The results reveal that Vwest reached 18.84 cm s−1 during December 2015, which is 57.1% and 14.3% faster than other winters in November 2012 (a neutral year, Vwest=11.99 cm s−1) and during November 2022, La Niña event (Vwest=16.48 cm s−1), respectively. Consequently, the stronger Kuroshio intrusion increases the westward currents in the Babuyan Channel and may further influence the ecosystem in the SCS, which are modulated by various phases of El Niño–Southern Oscillation (ENSO).