Skip to main content
. 2021 Nov 11;11:22027. doi: 10.1038/s41598-021-01189-3

Table 2.

The non-linear effect of different weather indicators at t-7 (at 12:00 local time) on different pandemic indicators.

(I) (II) (III)
(log) new cases New cases per 100 K capita (log) cases
Temperature spline < 25 °C − 0.0116*** (0.002) − 2.1171*** (0.395) − 0.0183*** (0.003)
Temperature spline > 25 °C − 0.0081*** (0.002) − 0.0617 (0.505) − 0.0089*** (0.003)
Humidity − 0.0012*** (0.000) − 0.1541*** (0.060) − 0.0016*** (0.000)
Wind speed − 0.0052*** (0.001) − 0.1515 (0.336) − 0.0074*** (0.002)
Precipitation 0.0244 (0.329) − 3.2391 (71.118) − 0.5390** (0.317)
Constant 1.2745*** (0.050) 147.0458*** (11.058) 4.2514*** (0.077)
Observations 595,251 595,251 595,251
R-squared 0.73 0.43 0.94
County fixed effect YES YES YES
State-date fixed effect YES YES YES

The outcome variables are (log) new cases (I), the number of new cases per 100,000 habitants within the last 14 days (II) and (log) cases (III). Standard deviations based on robust standard errors clustered at the county level in parentheses.

***,**,*Significance at 1, 5 and 10%, respectively; N = 595,251.