Abstract: To clarify the changes of atmospheric environment in Ningxia since 2013, PM₁₀ concentration over Ningxia from 2013 to 2016 was retrieved using a geographically weighted regression procedure based on satellite remote sensing, meteorological data and ground-based PM₁₀observation data. Results showed an obvious spatial variation of PM₁₀ concentration in Ningxia, with much higher in the middle semiarid regions and along Yellow River areas and much lower in the southern and northwest mountain regions. The yearly averaged PM₁₀ concentration from 2013 to 2016 over whole Ningxia ranged from 76 to 127 μg·m^-3, with a mean value of 100 μg·m^-3. There were seasonal differences of PM₁₀ concentration, with high values in winter (129 μg·m^-3) and spring (116 μg·m^-3) and low values in autumn (92 μg·m^-3) and summer (68 μg·m^-3). The interannual variation of PM₁₀ concentrations over Ningxia showed a decrease trend from 2013 to 2016, with a decrease rate of 15.8 μg·m^-3 per year. However, the differences of PM₁₀ concentration among different regions of Ningxia had decreased as indicated by the decreases of spatial standard deviation from 14.8 μg·m^-3 in 2013 to 6.5 μg·m^-3 in 2016, with a decrease rate of 2.8 μg·m^-3 per year.

Key words: mingling state, spatial structure, bivariate distribution., primary mixed broadleaved Korean pine (Pinus koraiensis) forest