Attributing Tropical Cyclogenesis to Equatorial Waves in the Western North Pacific

Abstract

Abstract Tropical cyclogenesis is attributed to an equatorial wave when the filtered rainfall anomaly exceeds a threshold value at the genesis location. It is argued that 0 mm day−1 (simply requiring a positive anomaly) is too small a threshold because unrelated noise can produce a positive anomaly. A threshold of 6 mm day−1 is too large because two-thirds of storms would have no precursor disturbance. Between these extremes, consistent results are found for a range of thresholds from 2 to 4 mm day−1. Roughly twice as many tropical cyclones are attributed to tropical depression (TD)-type disturbances as to equatorial Rossby waves, mixed Rossby–gravity waves, or Kelvin waves. The influence of the Madden–Julian oscillation (MJO) is even smaller. The use of variables such as vorticity and vertical wind shear in other studies gives a larger contribution for the MJO. It is suggested that its direct influence on the rainfall in forming tropical cyclones is less than for other variables. The impacts of tropical cyclone–related precipitation anomalies are also presented. Tropical cyclones can contribute more than 20% of the warm-season rainfall and 50% of its total variance. The influence of tropical cyclones on the equatorial wave spectrum is generally small. The exception occurs in shorter-wavelength westward-propagating waves, for which tropical cyclones represent up to 27% of the variance. Tropical cyclones also significantly contaminate wave-filtered rainfall anomalies in their immediate vicinity. To mitigate this effect, the tropical cyclone–related anomalies were removed before filtering in this study.