CD Tesis
Pemodelan Kelebatan dan Dura- si Hujan Skala Singkat (Setiap Jam) dengan Menggunakan Analisis Storm
Global climate change can lead to extreme weather events that often lead to losses. One of these extreme weather events is the storm rainfall. Therefore, it is necessary to model rainfall that can describe the characteristics of rainfall data as well as predict the same extreme events in the future. The research involves the modeling of storm rainfall, with data obtained using storm analysis techniques on short-scale (hourly) rainfall data obtained from the Department of Irrigation and Drainage Malaysia. The data represents rainfall over a span of 38 years (1970-2008) in Alor Setar, Malaysia.
There were 329,469 hourly rainfall data for 38 years analysed and from these data were taken data that met the event criteria storm. Defined events storm in this study as consecutive rainfall and no rainfall events that occurred over a duration of at least 3 hours, minimum width of 3 mm and the time interval between events storm for 3 hours with variable of the year’s maximum length and duration. Furthermore, the Generalized Extreme Value (GEV) and Generalized Pareto (GP) distributions are used to model rainfall data. There were four models produced, namely the GEVp and GPp models for the annual maximum rainfall frequency variables and the Gevt and the GPt model for the maximum annual rainfall duration variables. The four models were compared based on their variables using density graphs as well as AIC and BIC matching tests. This study showed that the best models to describe storm rainfall data in Alor Setar were the GEVp and GEV t models. Then obtained the return period of the most extreme rainfall with a maximum annual flexibility of 300 mm and maximum annual duration occurring over a period of about 40 years.
Keywords: GEV distribution, GP distribution, return period, short-term ra- infall, storm analysis.
Tidak tersedia versi lain