APLIKASI METODE MOMEN MOMEN PROBABILITAS TERBOBOTI UNTUK ESTIMASI PARAMETER DISTRIBUSI PARETO TERAMPAT PADA DATA CURAH HUJAN (Studi Kasus Data Curah Hujan Kota Semarang Tahun 2004-2013)

Abstract

The method used to analyze the extreme rainfall is Extreme Value Theory (EVT). One of the approaches in the EVT is Peak Over Threshold (POT) which follows the Generalized Pareto Distribution (GPD). The shape and scale parameter estimates obtained using the method of probability weighted moment. The results of this research were persumptive of maximum value in 1 period year in 2004 up to 2013 showed that in 2009/2010 has the biggest possibility of extreme value compared with other years. It’s also obtained the value of Mean Absolute Percentage Error values (MAPE ) 33,19 % . It is a big difference because the MAPE value is above 10%, thus it’s possible that extreme value appears Keywords: Rainfall, Extreme Value Theory, Over Peak Threshold, Generalized Pareto Distribution, Probability Weighted Moment