The Application of VECM (Vector Error Correction Model) for Rainfall Modeling in Manokwari Regency

Abstract

This study aims to determine the modeling of rainfall in Manokwari Regency with a Vector Error Correction Model (VECM) based on the period January 2012 to December 2017. In the modeling used temperature, humidity, and solar irradiation models. The research begins by looking at data stationary, if the data is not stationary, a differencing stage will be carried out to station the data. Then an optimal lag is determined and then a cointegration test is performed to determine whether there is a long-term relationship between variables on rainfall as a condition to proceed to VECM modeling. Based on VECM (1) obtained there are two relationships to changes in rainfall, namely the long-term and short-term relationships. In the long term relationship, changes in rainfall are only influenced by changes in the irradiation period of 1 (one) previous period. In the long term relationship, changes in rainfall are influenced by rainfall and irradiation period of 1 (one) previous period. In the short term relationship, changes in rainfall are only affected by changes in the irradiation period of 1 (one) previous period. When compared to ARIMA (1, 0, 2), VECM (1) can predict rainfall better because it has validation values smaller models than ARIMA (1, 0, 2).