Air‐sea coupled modes on intraseasonal and interannual time scales over the tropical western Pacific

Abstract

Dominant phase relationships between sea surface temperature (SST), high cloud cover (HCC), and zonal wind at 850 mbar on intraseasonal and interannual time scales in the warm pool region of the western Pacific are examined. Using the complex empirical orthogonal function (CEOF) analysis, two air‐sea coupled modes, that is, an intraseasonal mode and an interannual mode, are statistically extracted. It is found that in the intraseasonal mode the area of maximum SST anomaly lies to the east of the center of the active large‐scale convection area and therefore positive SST anomalies exist at the easterly anomaly area. The existence of high SST at the easterly anomaly area provides a favorable condition for a large‐scale disturbance to propagate eastward, though the evaporation‐wind feedback mechanism may also favor its eastward propagation if the mean wind is easterly. It is seen that the intraseasonal mode is very similar to the advective mode proposed by Lau and Shen (1988) with respect to the phase relationship. However, since the time evolution of the intraseasonal mode does not always coincide with the variations of east‐west SST gradient over the tropical western Pacific, this mode requires another mechanism other than the east‐west SST advection causing the advective mode. On one hand, the interannual mode has a characteristic that the phase relation of 850 mbar zonal wind (u) is different in each region of the tropical western Pacific, whereas the SST is almost in phase with HCC. The predominant growth of the mode defined in the key region (0°–10°N, 130°–160°E) corresponds to the life cycle that occurs when the El Niño/Southern Oscillation (ENSO) begins, develops, and decays. This verifies that the ENSO is one of the prominent phenomena in the air‐sea interactive system on the interannual time scale. On the other hand, the mode over the tropical Pacific around the dateline explains the air‐sea coupling when large‐scale convection is not very active. It is thus clear that evident differences of air‐sea coupling exist between the intraseasonal time scale and the interannual time scale. It is also determined that the strength of air‐sea coupling varies with year in both time scales. The above results provide a clue for understanding the mechanism of multiscale air‐sea coupling in the tropics.