Decadal variations of intense typhoon occurrence in the western North Pacific

Abstract

The causes of one of the two major oscillation periods, 16–32 years, identified through a wavelet analysis, of the time series of the frequency of intense typhoon (categories 4 and 5 in the Saffir–Simpson scale) occurrence for the period 1960–2005 in the western North Pacific (WNP) is studied in this paper. By dividing this period into sub-periods during which the frequency of intense typhoon occurrence was above or below normal on this time scale, various thermodynamic and dynamic factors in each sub-period are examined. During the above-normal periods, the sea surface temperature in the southeastern part of the WNP (5–20° N, 150–180° E) is found to be slightly higher. Within this area, the moist static energy (MSE) is also higher and the vertical gradient of saturated MSE in the lower troposphere is more negative. At the same time, the low-level streamfunction anomalies tend to have a negative maximum and the vertical wind shear between 200 and 850 hPa is also relatively small. Thus, both the thermodynamic and dynamic conditions within this area are more conducive to the development of tropical cyclones (TCs). As these cyclones move northwestward, the favourable dynamic conditions continue to be present so that they can intensify further. The steering flow is such that many of these typhoons will stay over water for an extended period of time through low-latitude recurvature. As a result, they can intensify to become category 4 or 5 typhoons. The conditions during the below-normal periods are generally opposite. A major conclusion from the results of this study is that the frequency of intense typhoon occurrence undergoes a strong multi-decadal (16–32 years) variation due to similar variations in the planetary scale oceanographic and atmospheric conditions that govern the formation, intensification and movement of TCs. These latter variations are largely contributed by the El Niño and the Pacific Decadal Oscillation on similar time scales.