Fig 1.
Exemplar high-resolution city-wise human mobility tracks and potential silent transmission of COVID-19 via co-location visits.
(A) The map and the distribution of population in Shijiazhuang, and the city center is enlarged and the exemplar mobility tracks from two individuals are shown on the right. (B) Their corresponding sequence of visited locations, numbered according to the location labels in the enlarged map, with the category of each location shown. Blocks in gray correspond to the period when the individuals are “moving” Their co-location visits, i.e. they stop at the same location in the same time window, are marked by dashed squares. The icons used in this figure were obtained or modified from open-source resources in Openclipart (https://openclipart.org/). The map was drawn based on open-source shape file with License Creative Commons BY 40 (CC BY 4.0) from OpenStreetMap (https://www.openstreetmap.org/).
Fig 2.
The spatiotemporal patterns of city-wise COVID-19 infection.
(A) The number of new infections ΔI(t) in the city as a function of time t at a 15-minute interval, given 70 randomly selected initial spreaders, averaged over 1000 realizations. A significant periodic pattern is observed, which is caused by the periodic human mobility behavior. Inset: The corresponding fraction of infected population, i.e. I(t)/N. (B) The distribution of the initial and the final infected population over the districts, i.e. Id(0)/I(0) and Id(T)/I(T) respectively; districts are numbered as shown in the map in C. (C) The evolution of the daily spatial pattern of the infected population in the city; the number of infected population in a location is represented by the color of the dot. The map was drawn based on open-source shape file with License Creative Commons BY 40 (CC BY 4.0) from OpenStreetMap (https://www.openstreetmap.org/).
Fig 3.
Infected professions and locations.
(A) The number of newly infected individuals of different professions p as a function of time t, i.e. ΔIp(t). Inset: the fraction of infected population from profession p who get infected at time t, i.e. ΔIp(t)/Ip(T), averaged over 1000 realizations. Periodic patterns still exist for different professions, but their infected population is largely different. (B) The number of transmissions in locations of different location category l as a function of time t, i.e. ΔIl(t). Inset: the fraction of transmissions in location category l which occur at time t, i.e. ΔIl(t)/Il(T). (C) Upper panel: the distribution of population over professions (orange bars), i.e. Np/N, which is proportional to the initial distribution of the infected population over professions Ip(0)/I(0), and the corresponding final infected distribution (green bars), i.e. Ip(T)/I(T); lower panel: the fraction of final infected individuals normalized by the population size in each profession, i.e. Ip(T)/Np. (D) Upper panel: the distribution of locations over location categories, i.e. Ml/M and the final distribution of transmissions over location categories, i.e. Il(T)/I(T); lower panel: the average number of transmissions in a single location of each location category, i.e. Il(T)/Ml.
Fig 4.
Transmissions inside and outside home and their dependence on age.
(A) The number of new transmissions inside and outside home as a function of time t, averaged over 1000 realizations. Inset: the fraction of transmissions inside and outside home normalized by the total number of transmissions as a function of time t, i.e. ΔI(t|home)/I(T|home) and ΔI(t|outside)/I(T|outside) respectively. (B) The number of new infected individuals as a function of time t in different age groups, i.e. ΔIa(t). Inset: the fraction of infections in different age groups which occur at time t, i.e. ΔIa(t)/Ia(T). (C) The distribution of initial and final infected population across different age groups, i.e. Ia(0)/I(0) and Ia(T)/I(T) as orange and green bars respectively, and the fraction of individuals infected at home in different age groups, i.e. Ia(T|home)/Ia(T) (red bars). (D) The fraction of population infected inside and outside their home according to their professions, i.e. Ip(T|home)/Ip(T) and Ip(T|outside)/Ip(T) respectively.
Fig 5.
The number of the infected population over professions until (A) the 3rd and (B) the 7th day of the simulations respectively, i.e. Ip(3rd day) and Ip(7th day), given that each of the initial infected group falls completely in four different professions. Similar investigation on (C) Il(3rd day) and (D) Il(7th day) for location categories, given that the initial infection starts at locations in four different location categories.
Table 1.
The average, median, 25th and 15th percentile of daily total traveling distance and the radius of gyration estimated from the mobility tracks of all users.
Table 2.
The statistics of the number of members per household in Shijiazhuang based on the 7th Census in 2022.