Studi Pemodelan Curah hujan sintetik dari beberapa stasiun di wilayah Pringsewu

Abstract

This research conducted to study the characteristics of daily rainfall and model making of
synthetic daily rainfall in Pringsewu regency using periodic model, stochastic model and periodic
stochastic models. This research conducted using daily rainfall data with length of 1984-2013
from three rainfall stations, Pringsewu, Wonokriyo and Banyuwangi rainfall stations.
These models performed by using 512 days annual data. Using rainfall frequency obtained and
applying the spectral method and the least squares method, it can be generated the daily rainfall
periodic models. Rainfall stochastic model assumed as the difference between rainfall data with
periodic rainfall models. Based on data from the series of stochastic, the component was
calculated using the approach of autoregressive models. Stochastic model was presented by using
the autoregressive model of order three. Periodic stochastic model obtained by merging periodic
model and stochastic model. Model validation and data obtained by calculating the correlation
coefficient. Based on the results of this research, it can be concluded that daily rainfall time series
can be very significantly approximately recorded rainfall data. With the the average value of
coefficient correlation of periodic model is 0.98019, coefficient correlation of stochastic model is
0.99808, and coefficient correlation stochastic of periodic model is 0.99993
keywords: daily rainfall, autoregresif models, stochastic component.