Fig. 7

The estimated counterfactual national cumulative confirmed cases by lockdown in \(r_4\) or \(r_5\) . We consider the baseline scenario as the epidemic curve estimated by S-SEIR model for A, B, C, D. A The estimated counterfactual national cumulative confirmed cases by lockdown in \(r_4\), which means the human mobility between some regions and \(r_4\) is set to 0, including travel flow from \(r_4\) (denoted as ’O’) and travel flow to \(r_4\) (denoted as ’D’). (B) The estimated counterfactual national cumulative confirmed cases by lockdown the flow in \(r_5\). (C) The median differences in estimated national cumulative confirmed cases at \(t = 30\) week by the destination-based lockdown in \(r_4\) and baseline scenario. (D) The median differences in estimated national cumulative confirmed cases at \(t = 30\) week by the destination-based lockdown in \(r_5\) and baseline scenarios. The size of the circle indicates the absolute value of the differences. Bottle green indicates that the cases estimated by counterfactual scenarios are fewer than the baseline. Purple indicates the cases estimated by counterfactual scenarios are higher. All distributions are obtained from \(n = 100\) ensemble members