出版社：The International Institute for Science, Technology and Education (IISTE)
摘要：Trends and patterns of extreme rainfall over Java Island, Indonesia were studied here for the period 1981-2010. The 9 low land and 10 mountainous gauges in respect to northern plain and southern mountainous area were selected after passing criteria such as gross error check, the restriction of at least 90% completed series, and homogeneity test. Some of recommended WMO/CLIVAR extreme indices were calculated annually and seasonally to express the frequency and its intensity. The total rainfall trend (PRCP), the frequency of heavy (R20mm) and very heavy (R50mm), and the simple daily intensity index (SDII) have been analyzed. The significance temporal trend was further assessed using the non-parametric Mann-Kendal test. The result showed that the extreme rainfall event over Java Island is depicted by an incoherent pattern which is spatially distributed across the island and is proportional between positive and negative trend with statistically not significant trend as a dominant changing. From annual extreme indices, it has been revealed that Banyuwangi, Malang, Tuban, Cilacap, and Jakarta Halim and Priok show a consistent increasing trend, while Jember, Probolinggo, Tegal, Semarang, Solo, Cirebon, Ciamis, and Serang show a consistent decreasing trend. Seasonal analyses exhibited a prominent negative trend during peak of dry season (JJA) with high coherency among nineteen stations and for some cases these decreases are followed by increasing SON rainfall indices which generate a hypothesis of a shifted rainfall from JJA to SON. Further analyses of seasonal indices have also detected some stations with a consistently increased DJF rainfall and decreased JJA rainfall e.g. Yogyakarta, Cilacap, Bogor, and Jakarta Priok which may lead a higher risk of hydro-meteorological hazard. Variability analyses on monthly rainfall series showed that the large scale phenomena such as ENSO may play an important role in the occurrence of extreme rainfall event.
