Open Access Open Access  Restricted Access Subscription Access

Time Series Analysis and Forecasting of Monthly Maximum Temperatures in South Eastern Nigeria


 

This study examines 1977 – 2012 monthly temperature data for South Eastern Nigeria collected using metrological instrument in the NRCRI Umudike at latitude 050, 291N and longitude 070, 331E (122M above sea level). A preliminary check on the time series plot of the data showed seasonal variation suggesting that the series was not stationary. The classical Box and Jenking’s Time Series methodology with its indicative ACF and PACF identification guide was employed. The SARIMA (0, 0, 2) (2, 1, 1)12 model was found to be adequate for the series and  monthly forecast from 2013 to 2017 showcased  relatively stable temperature values within these years.  Verification of the model using the 2011 – 2012 monthly data shows that the model is parsimoniously equitable. At the end, it was recommended that studying and carefully applying mathematical models could help track future rise in monthly temperatures.


Keywords

Temperature, Time Series, Seasonal ARIMA, ACF, PACF, Forecasting
User
Notifications
Font Size

Abstract Views: 156

PDF Views: 0




  • Time Series Analysis and Forecasting of Monthly Maximum Temperatures in South Eastern Nigeria

Abstract Views: 156  |  PDF Views: 0

Authors

Abstract


This study examines 1977 – 2012 monthly temperature data for South Eastern Nigeria collected using metrological instrument in the NRCRI Umudike at latitude 050, 291N and longitude 070, 331E (122M above sea level). A preliminary check on the time series plot of the data showed seasonal variation suggesting that the series was not stationary. The classical Box and Jenking’s Time Series methodology with its indicative ACF and PACF identification guide was employed. The SARIMA (0, 0, 2) (2, 1, 1)12 model was found to be adequate for the series and  monthly forecast from 2013 to 2017 showcased  relatively stable temperature values within these years.  Verification of the model using the 2011 – 2012 monthly data shows that the model is parsimoniously equitable. At the end, it was recommended that studying and carefully applying mathematical models could help track future rise in monthly temperatures.


Keywords


Temperature, Time Series, Seasonal ARIMA, ACF, PACF, Forecasting