Prediction of Global Warming Trends Using ARIMA-BP Neural Network and Their Implications for Water Resources

This paper quantitatively analyzes global warming using collected data and develops four predictive models: the Grey prediction model, ARIMA model, BP Neural Network model, and ARIMA-BP Neural Network model. The analysis reveals that the ARIMA-BP Neural Network model offers the best fit with minimal error, making it the most reliable for predicting future temperatures. Projections indicate that the Earth’s average surface temperature will not reach 20°C until the 22nd century, with predictions varying slightly among the models. Additionally, correlation analysis shows that global temperature is influenced by both time and location, with temperatures decreasing with latitude and stabilizing over time before rising due to global warming. The study also examines the regional impacts of natural disasters, such as volcanic eruptions and forest fires, and the global effect of COVID-19 on temperature. Based on these findings, the paper proposes measures to address global warming and mitigate its effects.