
Model 1

Model 2

Model 3


log(N_{im})

0.72^{∗∗∗} (0.69, 0.74)

0.48^{∗∗∗} (0.45, 0.51)

0.49^{∗∗∗} (0.46, 0.52)

log(f_{20}°)

2.18^{∗∗∗} (1.89, 2.48)

3.40^{∗∗∗} (3.01, 3.78)
 
log(GDP)

− 13.29^{∗∗∗} (− 14.20, − 12.38)
  
R squared

74.9%

57.9%

47.5%

 In model 1, all three explanatory variables log(N_{im}), log(f_{20}°), and log(GDP) are included. Model 2 includes log(N_{im}) and log(f_{20}° ). Model 3 only includes log(N_{im}). For each model, we report the regression coefficient (95% CI) for each of the explanatory variables along with R squared, based on n = 1220 cells
 n_{tr} average number of local ZIKV transmissions within each 0.25° × 0.25° cell from January 1, 2015, to December 31, 2016, N_{im} number of ZIKV importations, f_{20}° fraction of days with temperature higher than 20 °C, GDP gross domestic product per capita in purchasing power parity
 ***p < 0.001