Title: Forecasting CO2 emission of Turkey: swarm intelligence approaches
Authors: Eren Özceylan
Addresses: Department of Industrial Engineering, University of Gaziantep, 27310 Gaziantep, Turkey
Abstract: It is known that among the various greenhouse gases, CO2 is the most frequently implicated in global warming and climate change. Therefore, there is a need for developing efficient quantitative tools that allow forecasting CO2 emissions. This paper presents application of particle swarm optimisation (PSO) and artificial bee colony (ABC) techniques to estimate CO2 emission in Turkey, based on socio-economic indicators. The models are developed in three forms which are linear, exponential and quadratic. PSOCO2 and ABCCO2 (PSO and ABC CO2 estimation models) are developed to estimate the future CO2 emission values based on energy consumption, population, gross domestic product (GDP), and number of motor vehicles data. Emitted CO2 emission in Turkey from 1980 to 2008 is considered as the case of this study. While the first 25 years data of 29 years data is used for validation of four models, full data is used for future projections. Sum square error (SSE) is used as a fitness function in proposed models. Finally, CO2 emission in Turkey is forecasted up to year 2030 under different scenarios.
Keywords: artificial bee colony; ABC; CO2; carbon dioxide; carbon emissions; emission forecasting; particle swarm optimisation; PSO; Turkey; swarm intelligence; socio-economic indicators; energy consumption; population; gross domestic product; GDP; motor vehicles; sum square error; SSE.
International Journal of Global Warming, 2016 Vol.9 No.3, pp.337 - 361
Received: 08 Sep 2014
Accepted: 01 Nov 2014
Published online: 23 Mar 2016 *