Research on Climate Factors Based on DBSCAN Clustering Algorithm and ARIMA Time Series

Conducting scientific analysis and precise prediction of the time series characteristics of climate change is of vital theoretical and practical significance for understanding the laws of climate evolution and formulating response strategies. The temperature data of representative cities around the world are clustered through the DBSCAN algorithm, and temperature mutation tests and least squares regression calculations are conducted for each type of city to analyze and reveal the global temperature change trend. By constructing multiple time series models such as ARIMA, global temperature data is trained and predicted. Finally, through Spearman correlation analysis, the main factors influencing temperature were determined and a scientific basis was provided for formulating climate change mitigation measures.