Table 2.

Estimates of $\mathit{\alpha }$ and $\mathit{\beta }$ parameters (and corresponding p-values) for the ITS models implemented for NO${}_2$ time series. If the information is missing it means that the corresponding coefficient was not significantly different from zero and was removed from the model. Besides the covariate coefficients $\mathit{\beta }$, the following parameters are estimated: the intercept ${\alpha }_0$, the baseline trend slope ${\alpha }_1$ and the post-intervention level (trend slope) change ${\alpha }_2$ (${\alpha }_3$); see Eq. (1). In particular, ${\alpha }_0+{\alpha }_2+X_t\mathit{\beta }$ and ${\alpha }_1+{\alpha }_3$ represent the level and the trend slope after the lockdown, to be compared with ${\alpha }_0+X_t\mathit{\beta }$ and ${\alpha }_1$, respectively.

	Sensor ID
	6761		6781		30163
	Est.	$p$-value	Est.	$p$-value	Est.	$p$-value
${\alpha }_0$	62.84	$<$0.001	78.48	$<$0.001	61.45	$<$0.001
${\alpha }_1$	−0.21	0.005	−0.25	$<$0.001	−0.25	0.003
${\alpha }_2$	54.47	0.097	94.64	$<$0.001	43.52	0.200
${\alpha }_3$	−0.83	0.057	−1.42	$<$0.001	−0.68	0.138
${\beta }_{temperature}$
${\beta }_{precipitation}$			−0.66	0.018
${\beta }_{windspeed}$	−6.91	$<$0.001	−5.37	$<$0.001	−5.52	$<$0.001
${\beta }_{sunday}$	−8.44	$<$0.001	−5.56	$<$0.001	−7.24	$<$0.001