馬來半島極端降水之空間分布與長期變化

馬來半島位處於海－氣－陸交互作用相對複雜的海洋大陸地區，極端降雨的發生行為可說是不同尺度作用相互影響下所展現的結果。近年來，極端降雨事件在全球暖化及社會發展的衝擊下，對馬來半島各地已造成各方面嚴重的破壞及損失。因此，本研究將對馬來半島地區之極端降雨頻率和強度的時空變化進行更全面的探討，做為未來在災害潛勢評詁及都市發展等各方面的規劃，可因應極端降雨所帶來的天氣災害，並有效地降低災害的影響範圍。　　本研究使用馬來西亞水利灌溉局22個測站，時序從1971年1月至2007年12月共37年的日降雨資料進行分析。首先，本研究以小波時頻分析釐清馬來半島地區降雨週期的頻率變化，做為極端降雨分析的前置作業。在極端降雨的分析上，分別以超閾值法和最大值法做為極端降雨事件閾值之確定，分析極端降雨的長期變化趨勢及空間分布特性，並透過極值理論建立馬來半島極端降雨的概率分布模式，推估不同時期極端降雨的重現期，最後，透過極端降雨的個案分析，探討馬來半島地區極端降雨發生時的綜觀環境特徵。　　根據研究顯示，馬來半島東部地區的年最大降雨強度居半島之冠。春夏二季最大降雨強度主要分布在半島西部，秋冬季節最大降雨強度則分布在東部地區。極端降雨日由東往西遞減。以極端降雨趨勢的空間變化而言，在南北空間分布上，4°N以南地區的極端降雨發生頻率和強度都有增加的趨勢。另一方面，根據廣義極值分布(GEV)模式結果顯示，此分布模式對於半島西部地區極端降雨的重現水平估計難以掌握，就另一個觀點來看，這樣的結果也可能暗示了近年馬來半島極端降雨的趨勢在東西空間分布上的轉變。　　馬來半島極端降雨事件發生時的綜觀環境風場主要受到赤道西風、赤道東風和東北季風的影響，上述各環境風場在複雜地形的交互作用下形塑了馬來半島地區極端降雨的空間分布。在極端降雨空間分布的塑造上，半島西岸地區極端降雨的發生易受跨赤道流的影響；半島東岸地區的極端降雨主要分布在東北季風時期，而極端降雨的發生在南北分布上則受北近赤道槽的影響。此外，ENSO (El-Niño Southern Oscillation)事件對半島地區極端降雨的發生，離不開與季風環流的交互作用，而La-Niña現象對極端降雨的發生在於降雨強度的加劇。

關鍵字

極端降雨；小波時頻分析；廣義極值分布(GEV) ；近赤道槽；跨赤道流

並列摘要

Malaysia Peninsula is situated in the Maritime Continent with complex ocean-atmosphere-terrain interaction. The extreme rainfall behavior is regarded as a result of this complex multi-scale interaction. In recent years, the extreme rainfall event, aggravated by global warming and social development, has resulted in serious devastation and losses in Malaysia Peninsula. This study is expected to investigate the spatial-temporal variability of the frequency and intensity of extreme rainfall in Peninsula in a more comprehensive way. The finding of this study can be used for risk assessment and the urban development planning to reduce the impact of climate-related disaster caused by extreme rainfall in the future. 　　In this study, we made use of the daily rainfall data from a total of 22 rainfall stations obtained by the Department of Irrigation and Drainage Malaysia (DIDM) during January 1971 to December 2007. First, we have used the wavelets analysis to clarify the variability of rainfall periodicity in Malaysia Peninsula. We have employed peak over threshold (POT) and block maximum methods (BMM) to define extreme value threshold for the analysis of the long-term trends and variability and the spatial distribution of the extreme rainfall. In addition, we have used generalized extreme value (GEV) distribution to model the occurrence probability in any given years of extreme rainfall. Finally, we attempted to clarify the generalization characteristics of extreme rainfall pattern via the diagnostic case studies of extreme rainfall events. 　　The results showed that the greatest rainfall intensity was found in the eastern part of Peninsular Malaysia. Seasonally, the greatest rainfall intensity was recorded in the western Peninsula during the spring and the summer and in the eastern Peninsula during the fall and winter season. Moreover, the average number of days with annual extreme rainfall decreased from the east to the west of Peninsular Malaysia. For the north-south spatial distribution, this study showed that there was an increased trend in frequency and intensity of extreme rainfall to the south of the latitude 4 north. We also noticed a limitation on the use of GEV distribution to estimate the return levels of extreme rainfall in the western Peninsula. However, the result also showed that there is a shift of spatial distribution of extreme rainfall tendency in recent years. 　　The spatial distribution of extreme rainfall was greatly influenced by the interaction between the terrain and the synoptic environment dominated by the equatorial westerly, the equatorial easterly, and the northeast. Furthermore, the extreme rainfall events that occur in the western Peninsula were subjected to the impact of cross-equatorial flow, while the eastern Peninsula’s events were found during the northeast monsoon season. The latitudinal distribution of the extreme rainfall pattern is suggested to be caused by the migration of the northern near-equatorial trough. On the other hand, the ENSO episodes associated with the extreme rainfall events were also linked to the monsoon. In addition, the influence of La-Niña on extreme rainfall events have been shown to be on the increased intensity of extreme rainfall.