Climatological characteristics and seasonal forecasting of tropical cyclones making landfall along the South China coast

This paper presents the important climatological features of the tropical cyclones making landfall along the South China coast and proposes a statistical scheme for the prediction of the annual number of such tropical cyclones. This number is found to have a large variation, which is mainly due to the occurrence or nonoccurrence of the El Nino - Southern Oscillation (ENSO) phenomenon. A strong El Nino event is found to reduce the number of landfalling tropical cyclones whereas more tropical cyclones tend to make landfall in years associated with La Nina events. Such variations are more prominent in some seasons. The late season ( October - November) activity is generally suppressed ( enhanced) in El Nino ( La Nina) years whereas the chance of a tropical cyclone striking the South China coast increases ( decreases) significantly in the early season ( May and June) after the mature phase of a La Nina ( El Nino) event. These anomalous activities are apparently linked to the ENSO-induced anomalies in the low- and midlevel large-scale circulation. Based on the ENSO-related indices such as the Nino-3.4 sea surface temperature anomaly and the equatorial Southern Oscillation index, a statistical prediction scheme for the annual number of such landfalling tropical cyclones by 1 April is developed using the projection - pursuit regression technique. This scheme provides a 40% skill improvement in root-mean-square error with respect to climatology. A real-time prediction made in 2001 gave reasonable results.