Forecasting and modeling of atmospheric methane concentration

Methane, among the other factors, and its increasing concentration/quantification rate in the air play a significant role in global warming and climate change. The purpose of this research is to forecast the methane concentration in the air for the coming next 10 years using probabilistic time series models, i.e., autoregressive integrated moving average model, self-existing threshold autoregressive model, and smooth logistic transition autoregressive model for the methane data of Pakistan and its two neighboring countries, i.e., China and India from 1970 to 2012. The results indicated that in 2013–2020 the total concentration of methane in the air increased in Pakistan with two other countries. The higher CH 4 emission was in China. CH 4 followed the trends as China>India>Pakistan. Our forecasting found that after 2012 methane’s concentration level in the air increased for all the selected regions. The performance of each model is evaluated by mean error, root mean square error, and mean percent error. The autoregressive integrated moving average model performed better than the other models for Pakistan and China, while for India logistic smooth transition autoregressive model performed better than the other models.