Skip to main content
. 2020 Aug 21;28(1):775–788. doi: 10.1007/s11356-020-10483-y

Table 5.

Robustness with different time windows

40 days 20 days 15 days 10 days 5 days
AQI

− 15.4615***

(4.9904)

− 14.3614***

(4.7725)

− 14.2988***

(4.9492)

− 13.3812**

(5.2049)

− 13.8809***

(5.2240)
PM2.5

− 8.7017***

(3.008)

− 7.9388***

(2.8343)

− 8.8087***

(2.9042)

− 7.6451**

(3.0088)

− 7.3297**

(2.9573)
PM10

− 14.6715***

(4.4582)

− 15.7398***

(4.2637)

− 18.0572***

(4.2383)

− 17.3999***

(4.4486)

− 19.5900***

(5.0325)
SO2

− 1.1731**

(0.4440)

− 1.3917**

(0.5658)

− 1.1360**

(0.4553)

− 1.3910***

(0.4923)

− 1.4901**

(0.6209)
CO

− 0.0347

(0.0378)

− 0.0394

(0.0381)

− 0.0634

(0.0393)

− 0.0447

(0.0409)

− 0.0514

(0.0467)
NO2

− 4.1509

(1.9096)

− 5.7463***

(2.0615)

− 6.1853***

(2.1363)

− 6.4972***

(2.2504)

− 6.9538***

(2.5833)
O3 − 9.8474(5.9884)

− 11.5953*

(6.1217)

− 11.9761*

(6.2305)

− 10.7539

(6.5352)

− 9.7130

(7.0134)

Note: *, **, and *** respectively indicate significance levels of 10%, 5%, 1%. The table shows the regression coefficient of the event when each variable in the first column issued as the explained variable. This table combines multiple tables and omit other variables to save space. The data of Chengdu is deleted here