| Kuroshio intrudes into the South China Sea(SCS)through the Luzon Strait(LS)and shows complex variability characteristics.Substantial studies have been carried on the factors,forms and effects of Kuroshio intrusion into the SCS.However,little studies have touched the daily variation of the Kuroshio axis near the LS to reveal some new features.Satellite remote sensing data provides conditions for studies on the daily variation of Kuroshio,but the complicated current field makes it a to identify the daily Kuroshio axis near the LS.Based on satellite altimeter data,an improved synthetic method is proposed to identify and classify the daily Kuroshio surface axis near the LS.Based on the satellite altimeter geostrophic current,9496 axes are identified of 26 years from 1993-2018.The transition characteristics of the daily Kuroshio axis type are studied,and the relationship between the axis type transition and the wind,mesoscale eddies,Luzon Strait Transition(LST)and some air-sea indices is investigated.The main conclusion are as follows:(1)An improved synthetic identification method based on streamlines and velocities is proposed to solve the challenge of identifying the daily Kuroshio axis near the LS.Based on the daily geostrophic velocity derived from satellite altimeter data,the daily Kuroshio surface axis near the LS and the Kuroshio branch into the南海are identified,and the axes are classified into five types based on the identification results and fuzzy c-mean clustering method:Leaping,Weak leaking,Strong leaking,Looping Atypical,where Weak leaking and Strong leaking are newly discovered classifications,both of which show different characteristics.Atypical comprise of very few axes that obviously differ from the classical axis types.Compared with the velocity-based method,the synthetic identification method overcomes the interference to the identification from the complex daily current field in the LS.The classification results coincide with the classical Kuroshio intrusion classification method base on indices.(2)The seasonal variation of the Kuroshio axis types,the sea surface height and their relationship with the local wind field are studied.The Kuroshio axis types varied seasonally.Leaping(Looping and Strong leaking)prevails in summer(winter).Weak leaking shows less seasonality and is relatively less in winter.The composite sea level anomaly(SLA)patterns of Leaping and Looping are almost opposite,while the patterns of Weak leaking and Strong leaking are similar to the former(latter).The seasonal variation of the Kuroshio axis type is influenced by the local wind,and the correlation coefficient between the monthly occurrence frequency of Strong leaking plus Looping and the local wind direction reaches 0.67;the synthetic images of the 10 m wind field and wind stress rotational anomaly have good correspondence with the SLA images.(3)The interannual variability and trend of the Kuroshio axis type,and their relationship with the Luzon Strait surface transport(surface LST)were investigated.The correlation coefficient between the frequency of Strong leaking plus Looping and LST is-0.64.The interdecadal variation of Leaping(Looping)type shows a positive(negative)trend,mainly associated with the decrease of Looping in winter.(4)The daily transition characteristics of the axis types are studied,and it is found that the number of transitions between any two types(e.g.,Leaping→Looping and Looping → Leaping)in a month is similar,which is called transition symmetry,indicating that the axis tends to transition alternately between two types in a period of time.Using the contour-based method to identify mesoscale eddies,it is found that the axis symmetry is correlated with nearby eddy activities.The transition rate in ascending order is Looping < Leaping < Strong leaking < Weak leaking < Atypical.Looping and Leaping are the two relatively stable types.The Kuroshio axis type transition order is derived: in summer(winter),the Kuroshio axis types tend to transition between Leaping and Weak leaking(Strong leaking and Looping),while in spring and autumn,the axis types tend to transition between Leaping,Weak leaking,Strong leaking,and Looping. |