Forecasting Model of Rice Production Using Weighted Rainfall Index in Subang, Karawang, and Indramayu Regency
No Thumbnail Available
Date
2012-12-12
Journal Title
Journal ISSN
Volume Title
Publisher
Balai Besar Penelitian dan Pengembangan Sumberdaya Lahan Pertanian
Abstract
Description
Various forcasting models of rice production have been developed to support national food security. The forecasting models of national production which use recently have been carried out by the BPS and have not include the climate factors. Whereas, the climate factors influenced the rice’s production. The aim of this research is to develop the harvest area model using independent variables : Weighted Rainfall Index (WRI), SeaSurface Temperature (SST) Nino 3.4, and Dipole Mode Index (DMI). The models which developed was based on BPS models which consist of 3 periods. There are period 1 (January-April), period 2 (May-August), period 3 (September-December). Furthermore, rice production forecasting is the multiplication of harvest area and yield per ha. Rice production forecasting in one year is sum of the 3 periods. The research location are pantura areas, namely Karawang, Subang, and Indramayu. The result of the research showed that the model performance by WRI for period 2 (May-August) is better than period 1 and period 3. The mean of error for harvest area forecasting for periode 1, 2, and 3 of WRI variable, respectively is 14, 13, and 47%. Based on model validation, harvest area models by independent variable using WRI, SST Nino 3.4, DMI and ratio of harvest area and standard area, relatively have the same performance. One of the reasons is correlation between SST Nino 3.4 and DMI withrainfall is high. Mean of error for rice’s production forecasting of WRI are 13, 15, and 49%, while SST Nino 3.4, DMI, ratio of harvest area and standard area are 29, 12, and 51%. The range of error rice production forecasting at second year are 10-11%.