Forecasting Carbon Dioxide Emission of Asian Countries Using ARIMA and Simple Exponential Smoothing Models

Abstract

Industrial revolution has increased the level of pollution due to emission of carbon dioxide in the past few decades. Resulting changes in the world’s climate are seen in the form of an increase in temperature which causes heat waves, diseases related to skin, cardiovascular and other respiratory illnesses in humans. According to a past study, overall contribution of continent Asia is 49% - 50% in the production of carbon dioxide. Therefore, data of carbon dioxideemission, heat and electricity, manufacturing industries, residential and commercial buildings, transport and other sources of Asian countries namely: Japan, Bangladesh, China, Pakistan, India, Sri Lanka, Iran, Singapore, and Nepal from 1971 to 2014 have been taken for the present study. Empirical analysis shows that heat and electricity arethe main cause to increase carbon dioxideemission in Pakistan, Bangladesh, India, Iran and Sri Lanka. Whereas this emission is increased in China due to residential and commercial buildings and transport variables. On the other hand, for Nepal and Singapore, these variables do not play any significant role in CO2 emission. Simple exponential smoothing (SES) and ARIMA models are used to forecast CO2 emissions. Furthermore, simple exponential smoothingis found appropriate for Pakistan and Sri Lanka based on the minimum FMAE. Whereas, for Japan, China, India, Iran and Singapore, ARIMA model is found suitable as it has minimum FMAE value. In Nepal and Bangladesh, both models have significant differences, so any one of them can be used for forecasting

Keywords

• simple exponential smoothing
• carbon dioxide (co2)
• multiple linear regression.