Abstract: In order to study the nonlinear relationship between PM_2.5 concentration variations and complex factors in Guanzhong region, the spatial distribution characteristics of PM_2.5 on different time scales such as year, season and month were deeply analyzed based on the meteorological data from January 2020 to December 2023. The maximum information coefficient was used to examine the relationship between PM_2.5 and other air pollutants in the region, and the CatBoost-SHAP-MCM model was applied to identify the key meteorological factors affecting PM_2.5 concentrations. The results show that: (1) The PM_2.5 concentration in the Guanzhong region exhibits obvious spatial distribution and seasonal variation characteristics, with the lowest annual average PM_2.5 concentration of 42.93 μg/m³ in 2021 and the highest of 49.09 μg/m³ in 2022. Seasonal and monthly variations are similar, both showing heavy pollution in winter and low pollution in summer. Pollution is most serious in winter, with PM_2.5 concentration reaching 84.35 μg/m³, and the lowest pollution in summer, at 21.42 μg/m³. Xi′an, Xianyang and Weinan are high-pollution cities, while Tongchuan and Baoji are low-pollution cities. (2) PM_2.5 shows the highest correlation with PM₁₀ concentrations, while correlations with CO and SO₂ concentrations are relatively low. (3) Dew point temperature, air temperature, sea level pressure, precipitation and surface pressure are the key meteorological influencing factors. Their influence is still significant in each city, and their effects are consistent across both the entire Guanzhong region and individual cities. (4) Under specific combinations such as low dew point temperature with low air temperature, and low dew point temperature with high sea level pressure and high surface pressure, the effects on PM_2.5 concentrations become more pronounced. The study demonstrates that PM_2.5 concentrations in the Guanzhong region exhibit distinct spatial differentiation and seasonal variation patterns, and are closely related to meteorological factors such as dew point temperature, air temperature, sea level pressure, precipitation, and surface pressure. These factors lead to more significant fluctuations in PM_2.5 concentration fluctuations under certain meteorological combinations.