Global Temperature Analysis Based on ARIMA Model and Pearson's Correlation Coefficient

This paper focuses on the global temperature trends and factors affecting the temperature by using regression model and ARIMA Model. This paper visualizes and analyzes the data by plotting a series of images and then concluds that there was no anomalous increase in global temperatures in March 2022. This paper builds a regression model and an ARIMA model to characterize past and predict future data. This paper solves for the average global temperature in 2020 using a quadratic polynomial regression model and an ARIMA model. This paper solves for a global average temperature of 20 degrees for the years 2159 and 2218, respectively. This paper finds that the different models have different results and perform differently in describing the past and predicting the future. Besides, this paper concludes that the ARIMA model is more suitable for describing existing data, forecasting the future and making short-term predictions, while the quadratic polynomial regression model is more suitable for predicting the future. Then this paper uses Pearson's correlation coefficient to simplify location to latitude, and then concludes that the higher the latitude, the lower the temperature. These three quantities are then put together to solve for the relationship when they work together. It was found that time and latitude have a very strong effect on temperature, while the effect of longitude is not significant. The results of the models used in this paper are all better and have some practical significance.