*3.3. Relationships between the Concentrations of Particulate Matter and Meteorological Factors*

Spearman correlation coefficients between PM10 (and PM2.5) and meteorological factors varied across seasons, indicating that the dominant meteorological factors and their influence on PM10 and PM2.5 vary over time (Tables 3 and 4).


**Table 3.** Spearman correlations and multivariate progressive linear regression between PM10 and meteorological factors in Xi'an.

\* indicates a significance level of 0.05; \*\* indicates a significance level of 0.01.

**Table 4.** Spearman correlations and multivariate progressive linear regression between PM2.5 and meteorological factors in Xi'an.


\* indicates a significance level of 0.05; \*\* indicates a significance level of 0.01.

Precipitation was negatively correlated with PM10 and PM2.5 (*p* < 0.01), except for PM2.5 in winter (*p* > 0.05). Atmospheric pressure was negatively correlated with PM10 and PM2.5 in summer and winter (*p* < 0.05), and positively correlated with the annual PM10 and PM2.5 (*p* < 0.05). Relative humidity showed a negative correlation (*p* < 0.05) with the annual and seasonal concentrations of PM10, except in winter. This same correlation with relative humidity was also exhibited for PM2.5 in spring and autumn, though a positive correlation was found with PM10 and PM2.5 in winter. Atmospheric temperature exhibited a negative correlation with PM10 and PM2.5 in spring and autumn, and a positive correlation in summer and winter. Wind speed only exhibited a negative correlation (*p* < 0.05) with PM10 and PM2.5 in autumn and winter. Finally, wind direction exhibited a negative correlation (*p* < 0.05) with PM10 and PM2.5 in all seasons except for autumn.

#### *3.4. Response of Atmospheric Pollutants to the Environmental Management*

Figure 4 displayed the emission intensity of SO2, smoke, and dust, and environmental regulation intensity in 2013–2017 in Xi'an. With the growth of economy, the emission intensity of SO2, smoke and dust exhibited declining trends during 2013–2017, especially SO2. In 2015–2016, the emission intensity of these atmospheric pollutants sharply declined from 9.49 to 1.20 t/108 RMB. However, the environmental regulation intensity (ERI) exhibited increasing trends during 2013–2017. In 2015–2016, ERI dramatically increased from 3.27 to 9.98.

ERI was significantly and negatively correlated with the emission intensity of SO2, smoke and dust (*p* < 0.05), but not significantly correlated with the concentration of PM10 and PM2.5 (*p* > 0.05).

**Figure 4.** Environmental regulation intensity and the emission intensity of SO2, smoke, and dust.
