The Daily and Hourly Rainfall Data Modeling Using Vector Autoregressive (VAR) with Maximum Likelihood Estimator (MLE) and Bayesian Method (Case Study in Sampean Watershed of Bondowoso, Indonesia)
Preprint
- 15 October 2018
- preprint
- Published by MDPI AG in Preprints.org
Abstract
The hourly and daily rainfall data which is spatially distributed are required as an input for run-off rain model. Furthermore, the run-off rain model is used to detect early flooding. The daily and hourly rainfall data have characteristics that most of data are zero. Therefore we need a model which can capture the phenomenon. A time series model involving location, which is a model that can be developed to approach the daily and hourly rainfall data, we can call Vector Autoregressive (VAR) model. The VAR model allows us for modeling rainfall data in several areas. However, in certain conditions the VAR model often occurs over-parameterization and reduces degrees of freedom. The aim of this study is to compare the VAR model with Maximum Likelihood Estimator (MLE) and Bayesian to hourly and daily rainfall data in SampeanWatershed of Bondowoso. The results showed that the hourly and daily rainfall data are fitted to VAR process of orde 5 and 1 respectively. Based on the AIC and SBC values indicate that the Bayesian is better than the MLE method. The Bayesian is able to predict parameters by producing a smaller variance covariance matrix than the MLE.Keywords
All Related Versions
This publication has 5 references indexed in Scilit:
- Analyzing and Forecasting Output Gap and Inflation using Bayesian Vector Auto Regression (BVAR) Method: A Case of PakistanInternational Journal of Economics and Finance, 2014
- Bayesian Multivariate Time Series Methods for Empirical MacroeconomicsFoundations and Trends® in Econometrics, 2010
- Bayesian Methods for Dynamic Multivariate ModelsInternational Economic Review, 1998
- NUMERICAL METHODS FOR ESTIMATION AND INFERENCE IN BAYESIAN VAR-MODELSJournal of Applied Econometrics, 1997
- Stastical Decision Theory and Bayesian Analysis.Journal of the American Statistical Association, 1988