PENDEKATAN REGRESI KOMPONEN UTAMA DAN ARIMA UNTUK STATISTICAL DOWNSCALING

Khairil Anwar Notodiputro, Aji Hamim Wigena, Fitriadi Fitriadi

Abstract


General Circulation Models (GCM) is a sophisticated computer simulation model concerning climate and its
components, such as weather temperature, water precipitation, as well as how these components change
according to time. GCM produces data in term of grid of an area with low resolution (2.50 or ± 300 km2)
reflecting global climate condition. Hence, these data are not measured in local or regional scale. Statistical
downscaling is a method useful to study climate change based on the GCM data. This statistical method
relates global and local climate variables as a projection of GCM output in local scale. However, since the
GCM output is basically a high dimensional time series data then standard statistical procedures would not be
appropriate. This paper demonstrates that the accuracy and precision of the statistical downscaling could be
improved through the use of principal component regression techniques in which the ARIMA models were
applied to the regression error.
Keywords: statistical downscaling, principal component analysis, ARIMA, regression analysis,
general circulation models

Article Metrics

Abstract view : 531 times
PDF - 1107 times

Full Text:

PDF

Refbacks

  • There are currently no refbacks.
Shell download

Shell download sitesi

Sunucularınız da deneme ve yanılma yöntemi ile birşeyler yapmak için shell download ederek hemen sunucularınızı test etmenizi tavsiye ediyoruz