Author(s)

Xiaolin Jin, Changquan Huang

Corresponding Author

Changquan Huang

Abstract

Global warming will cause little harm to the environment, which in turn will affect people's life and production. In this paper, we use time series models to predict the future global temperature change and explore the main causes of global warming through information gain. To address question 1, firstly, the monthly temperature from 2012-2021 was selected for a one-sample t-test with the temperature in March 2022, and it was obtained that the global temperature in March 2022 increased more than during the past decade. The ARIMA model and the LSTM model are used to predict the future global temperature level. For question 2, using the gray correlation analysis model, we found that there is a strong connection between global temperature and regions, and the southern hemisphere region has a greater influence on global temperature. Then, by establishing a multiple regression model to determine the relationship between global temperature and natural disaster factors. Finally, by calculating the indicators of carbon dioxide, forest area, and population size through information gain, it is concluded that the main cause of global warming is due to the excessive emission of carbon dioxide. In response to question three, in the future, we should reduce the use of energy sources such as coal and use more green and clean energy sources, thus reducing carbon dioxide emissions. Plant trees and increase forest cover, thus effectively curbing global warming.

Keywords

Global warming; One-sample t-test; ARIMA; LSTM; Information gain