The low-level cross equatorial flow over the South China Sea (SCS) can be separated to summer and winter regimes. The transition from the winter to the summer regime is associated with withdrawal of the stationary ridge of the Western Pacific Subtropical High from the SCS. In this study, the temporal and spatial characteristics of the seasonal transition of the low-level flow are investigated. A transition index is defined for the purpose of characterizing the onset of the transition. Using 43-year NCEP/NCAR Reanalysis data, we find the transition can be categorized into four groups, namely, early, normal-early, normal-late, and late. The inter-group differences in the aspect of regional process of the transition are small, but in large-scale anomalous patterns are distinct.

In the early transition year, the SST anomalies over the Philippine Sea and the far western Pacific are warmer than normal. The warm anomalies, originating from the warm SST over the SCS in the November of the previous year, are associated with abnormally strong southerly surface winds. The flow transition over the SCS is earlier than usual because the large-scale condition is favorable for strong low-level convergence over the Bay of Bengal, SCS and the surrounding maritime continents.

In the late transition years, the SST anomalies over the far western tropical Pacific are colder than normal. The cold anomalies can be traced back to the cold SST over the marginal seas near Australia and Indonesia in the October of the previous year. The cold SST anomalies are associated with abnormally low SLP over the eastern Pacific first over both Northern and Southern Hemispheres in boreal autumn and then only over the Southern Hemisphere in boreal winter. The low-level equatorial westerly winds in association with the anomalous large-scale SLP patterns are particularly strong. The delayed transition may be caused by such a condition that the large area from the Arabian Sea to the western Pacific is not favorable for low-level convergence owing to the abnormal upward motion over the South Eastern Pacific.

Although in the early and late groups the abnormal large-scale patterns can be identified at least one season before the transition, the patterns do not persist after the transition. In addition, no clear correlation between the transition timing and ENSO phases are found. However, after the cold phase in boreal winter, the possibility of a delayed transition is small.