Validity Investigation and Temperature Prediction of Global Warming Based on Random Forest Algorithm

This study initially examined the validity of the global warming argument through normal distribution test and one-sample t-test, and conducted a data perspective analysis to determine whether the increase of global temperature in March 2022 was larger than any previous 10-year period. Subsequently, based on the data collection and processing of global temperature impact factors and correlation analysis, three prediction models were established: multiple regression prediction model, ARIMA time series prediction model, and random forest-based nonlinear prediction model. These models collaborated to complete a forecast of future global temperatures. After accuracy assessment and sensitivity analysis with confidence intervals ranging from 95% to 105%, the nonlinear prediction model based on the Random Forest algorithm was selected to further study the future global temperature. The results showed that the global average temperature would reach 20 degrees in 2055, 19.96 degrees in 2050, and 20.07 degrees in 2100.