14:30 〜 14:45
[MIS06-04] Multiscale variability of extreme precipitation events in Taiwan based on weather types
キーワード：Taiwan, extreme precipitation, weather type
Taiwan, ranked as having the highest mortality risk, highest economic risk, and as being most exposed to “multiple hazards” than any other country by the World Bank (Dilley et al. 2005), is frequently faced with the threats caused by hydrological extremes resulting from excessive precipitation. The National Science and Technology Center for Disaster Reduction (NCDR) in Taiwan is responsible for monitoring the risks of the disastrous weathers to implement timely early warning actions. It is thus critical for NCDR to study the characteristics of extreme precipitation to mitigate impacts possibly brought by hydro-meteorological disasters.
Though the changes in Taiwan’s extreme precipitation characteristics have been investigated in a number of studies, how the diversity of weather types affected the extreme precipitation were less emphasized. In this study, we analyzed the seasonal distribution of Taiwan’s extreme precipitation occurrence by various weather types. We also analyzed the interannual variations of these extreme precipitation occurrences with respect to seasons. Dividing a year into five natural seasons: spring, Meiyu season, typhoon season, autumn and winter, we found that the season most prone to extreme precipitation events was the typhoon season, followed by the Meiyu season and autumn. Tropical cyclones, fronts, tropical cyclones accompanied by northeasterly monsoon (hereinafter referred as TC-NE), and diurnal/afternoon convection were the four major weather types that caused extreme precipitation events. In general, the highest proportion of extreme precipitation events was caused by the tropical cyclones. In terms of major weather types that caused extreme precipitation in different seasons, the Meiyu season (May-June) was dominated by the frontal type, and the typhoon season (July-September) was dominated by the tropical cyclone type, followed by the diurnal/afternoon convection type. In autumn (October-November), the TC-NE type was the majority. In terms of the interannual variability, during the analysis period from 1993 to 2015, the extreme precipitation occurrence in the typhoon season exhibited noticeable bi-decadal changes, showing an upward trend and then a downward trend. It was found that these changes were mainly contributed by the tropical cyclone type events.
Compared with the extreme precipitation in the Meiyu season and typhoon season, which have been more frequently studied in the past, there are fewer studies on the extreme precipitation in autumn. We aimed at autumn season to further investigate the climatic characteristics of the extreme precipitation. Based on 41 years of precipitation observations, and by relaxing the threshold of daily precipitation to obtain more event samples, we found that the TC-NE, the tropical cyclones, and the northeasterly are the three major weather types related to the extreme precipitation occurrence in autumn. The extreme precipitation occurrence showed significant variabilities in the decadal (10-11 year period) scale and interannual (2-7 year period) scale. Extreme precipitation events are more likely to occur in the autumn of the negative phase of the Pacific Decadal Oscillation and the autumn of the La Nina phase, respectively. Furthermore, three different weather types differed in their relationships with PDO and ENSO, respectively. The large-scale circulation characteristics responsible for these differences were discussed.
Though the changes in Taiwan’s extreme precipitation characteristics have been investigated in a number of studies, how the diversity of weather types affected the extreme precipitation were less emphasized. In this study, we analyzed the seasonal distribution of Taiwan’s extreme precipitation occurrence by various weather types. We also analyzed the interannual variations of these extreme precipitation occurrences with respect to seasons. Dividing a year into five natural seasons: spring, Meiyu season, typhoon season, autumn and winter, we found that the season most prone to extreme precipitation events was the typhoon season, followed by the Meiyu season and autumn. Tropical cyclones, fronts, tropical cyclones accompanied by northeasterly monsoon (hereinafter referred as TC-NE), and diurnal/afternoon convection were the four major weather types that caused extreme precipitation events. In general, the highest proportion of extreme precipitation events was caused by the tropical cyclones. In terms of major weather types that caused extreme precipitation in different seasons, the Meiyu season (May-June) was dominated by the frontal type, and the typhoon season (July-September) was dominated by the tropical cyclone type, followed by the diurnal/afternoon convection type. In autumn (October-November), the TC-NE type was the majority. In terms of the interannual variability, during the analysis period from 1993 to 2015, the extreme precipitation occurrence in the typhoon season exhibited noticeable bi-decadal changes, showing an upward trend and then a downward trend. It was found that these changes were mainly contributed by the tropical cyclone type events.
Compared with the extreme precipitation in the Meiyu season and typhoon season, which have been more frequently studied in the past, there are fewer studies on the extreme precipitation in autumn. We aimed at autumn season to further investigate the climatic characteristics of the extreme precipitation. Based on 41 years of precipitation observations, and by relaxing the threshold of daily precipitation to obtain more event samples, we found that the TC-NE, the tropical cyclones, and the northeasterly are the three major weather types related to the extreme precipitation occurrence in autumn. The extreme precipitation occurrence showed significant variabilities in the decadal (10-11 year period) scale and interannual (2-7 year period) scale. Extreme precipitation events are more likely to occur in the autumn of the negative phase of the Pacific Decadal Oscillation and the autumn of the La Nina phase, respectively. Furthermore, three different weather types differed in their relationships with PDO and ENSO, respectively. The large-scale circulation characteristics responsible for these differences were discussed.