1：Experimental Design and Method Selection:

The study uses frontal detection on satellite-derived SST images to determine the path of the Kuroshio Current and analyzes climate factors like ONI and PDO. An entropy-based edge detection method is applied for SST front determination without subjective thresholds, and path analysis is used to define causation between parameters.

2：Sample Selection and Data Sources:

Data includes monthly mean SST images from MODIS Aqua (2003-2016, 4km resolution) and OSTIA (1985-2016, 5km resolution), geostrophic current data from AVISO (1993-2016), and wind stress data from QuikSCAT (2000-2008) and ASCAT (2009-2016), all uniformed to 0.25°×0.25° resolution. Sampling areas are S (120.1-121.7°E, 19.6-21.2°N) and N (121.4-123°E, 25.2-26.9°N).

3：25°×25° resolution. Sampling areas are S (1-7°E, 6-2°N) and N (4-123°E, 2-9°N). List of Experimental Equipment and Materials:

3. List of Experimental Equipment and Materials: Satellite sensors (MODIS Aqua, Pathfinder AVHRR, ATSR, QuikSCAT, ASCAT), in-situ observations from ICOADS dataset, and computational tools for data processing and analysis.

4：Experimental Procedures and Operational Workflow:

SST images are processed using edge detection to compute gradient magnitudes (GM) for frontal pixels, monthly mean GM is calculated, and path analysis is performed on correlation matrices to model causal relationships between variables like GM, current velocity, wind stress, ONI, and PDO.

5：Data Analysis Methods:

Statistical analysis including correlation coefficients, time series trend analysis, and path analysis with standardized regression coefficients to determine significant relationships and trends.