Skip to main content
. 2017 May 22;14(5):549. doi: 10.3390/ijerph14050549

Table 1.

Variance explained for each predictive variable.

Predictive Variable (Unit) Variance Explained in the Univariate Model Variance Explained in the Multivariate Model (without PM10) Variance Explained in the Multivariate Model (Including PM10)
PM10 (μg/m3) 73.00% - 67.97%
Aerosol optical thickness (AOT) 7.38% 4.77% 0.48%
Normalized difference vegetation index (NDVI) 3.14% 0.24% 0.24%
Precipitation (kg/m2s) 1.75% 0.02% 0.02%
Temperature (°C) 1.08% 2.62% 0.48%
Mean specific humidity (kg/kg) 9.08% 0.48% 0.48%
Roadway length within the 10 km buffer of a monitoring station (m) 2.62% 2.86% 1.45%
Shortest distance of roadway to a monitoring station (m) 1.73% 2.15% 1.21%
Wind speed vector 3.76% 1.43% 0.48%
Area proportion of the factories and mines, oil fields and stone-pit land-use 2.29% 2.15% 0.73%
Area proportion of the forest land-use 2.06% 2.62% 0.48%
Number of the emission plants 4.48% 1.67% 0.97%
Shortest distance to the emission plants 1.70% 1.67% 0.24%
The first temporal basis function 37.00% 26.71% 4.84%
The second temporal basis function 5.79% 1.43% 0.48%
Time (day of year) 14.70% 2.38% 0.73%
Total 53.20% 81.30%