Time series modelling of monthly rainfall in southern Kerala

Authors

  • R. S. Neethu
  • Brigit Joseph

Abstract

This paper aimed to fit SARIMA model based on Box-Jenkins methodology to the time series data corresponds to monthly rainfall in three agro climatic regions viz., Regional Agricultural Research Station (RARS) Vellayani, RARS Kumarakom and Cardamom Research Station (CRS), Pampadumpara representing different regions of Southern part of Kerala. The empirical model gave a picture of climate change scenario happened in both temporal and regional wise. The SARIMA model was fitted to monthly rainfall for all the regions Vellayani, Kumarakom, and Pampadumpara using the data for 31 years from 1991 to 2021. The best identified SARIMA models for rainfall were ARIMA (1, 0, 0) (0, 1, 1)12, ARIMA (0, 0, 0) (0, 1, 1)12 and ARIMA (0, 0, 0) (0, 1, 1)12. The model parameters were obtained by using maximum likelihood method and the best model were selected using Akaike Information Criteria (AIC), Bayesian Information Criteria (BIC) and Hannan-quinn coefficient. The adequacy of the check of the selected models confirmed that the selected models were free from autocorrelation and the residuals are normally distributed.

Downloads

Published

2023-11-01

How to Cite

R. S. Neethu, & Brigit Joseph. (2023). Time series modelling of monthly rainfall in southern Kerala. JOURNAL OF ROOT CROPS, 47(1 & 2), 29–35. Retrieved from https://journal.isrc.in/index.php/jrc/article/view/620