Table 2. Statistics of the modelled air pollution concentrations and predictor variables for 8000 house address locations in the province of Utrecht resulting from the comparison of the vector-based and raster-based modelling approaches.

Variable (units)	Buffer size (m)	r2	RMSE	Bias
Units of RMSE and Bias are the same as the corresponding variable, their calculations according to Equations 1 and 2.
NO2 (μg m−3)	n/a	0.98	1.37	1.05
NOx (μg m−3)	n/a	0.98	1.9	1.13
PM2.5 (μg m−3)	n/a	0.99	0.034	6.3e-3
PM2.5absorbance (10−5 m−1)	n/a	0.98	0.043	2.9e-2
PM10 (μg m−3)	n/a	0.99	0.041	9.2e-3
TL (vehicles day−1 m)	500	0.95	1.24e7	8.43e6
RL (m)	1000	0.99	7.81e3	6773.61
HTL (vehicles day−1 m)	25	0.93	1731.53	396.77
	50	0.97	7111.19	2368.91
	500	0.98	5.94e5	−3.73e5
IND (m2)	5000	1.0	7.56e4	−7859.48
HAR (m2)	5000	1.0	3151.88	−42.71
POP (inhabitants m−2)	1000	1.0	14.25	−5.57
	5000	1.0	913.69	88.49
TML (vehicles day−1 m)	1000	0.99	3.89e6	1.59e6
RES (m2)	5000	0.99	2.42e5	−1.17e5
IDC (m−1)	n/a	0.85	0.026	4.3e-4
IDM (m−1)	n/a	0.91	4.1e-3	4.5e-5
BEO (μg m−3)	n/a	1.0	0.06	−0.06
BEX (μg m−3)	n/a	1.0	0.11	−0.1
BEP (μg m−3)	n/a	1.0	1e-4	5e-5
BEA (10−5 m−1)	n/a	1.0	4.64e-7	−7e-6