Skip to main content
. 2019 Apr 11;16(7):1300. doi: 10.3390/ijerph16071300

Table 2.

Land use regression model for annual average ozone concentration (ppb).

Variable Regression Coefficient p-Value Partial R
Intercept 1.52 <0.01
NOx −4.79 × 10−3 <0.01 0.54
Thermal power plant −1.55 × 10−6 <0.01 0.08
All types of residential—25 m −1.25 × 10−5 0.06 0.001
Relative humidity −1.85 × 10−3 <0.01 0.02
Forest—500 m 1.15 × 10−7 <0.01 0.02
Altitude 1.03 × 10−4 <0.01 0.009
Distance to main road 9.64 × 10−6 <0.01 0.005
Purely residential—25 m −3.25 × 10−6 0.13 0.004
Cemetery and crematorium—3000 m −1.71 × 10−8 <0.01 0.004
Temple—500 m −4.29 × 10−3 0.01 0.003
Temperature 9.05 × 10−3 <0.01 0.003
Non-irrigated crops—250 m 2.09 × 10−7 <0.01 0.002
Temple—1000 m −4.13 × 10−4 0.05 0.001
Mixed residential area—25 m −9.25 × 10−4 <0.01 0.04
Industrial area—5000 m −1.44 × 10−9 <0.01 0.003

Model performance: overall model R2 = 0.74; adjusted R2 = 0.73; Root Mean Square Error (RMSE) = 0.04 ppb; 10-fold cross-validation R2 = 0.70; externally validated R2 = 0.39.