Pulse fractional grey model application in forecasting global carbon emission

Global climate problem has attracted attention from all over the world. How to correctly reflect greenhouse gas emission trend is not only an environmental problem, but also concerns human society's sustainable development. Energy consumption is the primary source of greenhouse gas emission. It is necessary to accurately forecasting energy consumption greenhouse gas emission future trend. COVID-19 epidemic has brought respite from global climate change problem by reducing energy consumption through home office, work stoppage, and global travel ban. However, in the post-epidemic period, energy consumption greenhouse gas emission intensity trend has become the center research field. This study takes global energy consumption carbon emission problem as the main research line and focuses on forecasting global energy consumption carbon dioxide emission trend in major regions based on COVID-19 epidemic shock background. The study considering epidemic shock volatility characteristic. Firstly, fractional order grey mode (FGM(1,1))is used as the baseline model to balance time series data weight. Secondly, Median absolute deviation data preprocessing is introduced to reduce data fluctuation. Finally, a novel delaying pulse shock function optimized background grey forecasting model is also proposed to reflect epidemic shock-response characteristic. The proposed model is compared with existing models. It is found that data preprocessing and novel proposed model not only improves historical data's fitting quality by reflecting COVID-19 epidemic's shock characteristic, but also showed excellent forecasting performance for future trend. The novel grey model largely solves existing model underfitting/overfitting problem. In the end, based on forecasted results, we summarize research conclusion and implication.