Potential dissemination of epidemics based on Brazilian mobile geolocation data. Part I: Population dynamics and future spreading of infection in the states of Sao Paulo and Rio de Janeiro during the pandemic of COVID-19.

Pedro S. Peixoto; Diego R. Marcondes; Cláudia M Peixoto; Lucas Queiroz; Rafael Gouveia; Afonso Delgado; Sérgio M Oliva

The SI model is not suited for predicting the incidence of the infection for long periods of time, but it is an adequate linear approximation for the early exponential spread. The model chosen was adequate to be used with the available disease information, namely, the Basic Reproduction Number R 0 estimated from the initial spread in China. We also introduced , a free parameter, used to correct the overestimation or underestimation of movement between the locations. As expected, parameter is related to the intensity of mobility, which in turn implies a greater or smaller time of infection for each city. This is an indicative that the decrease in mobility, enforced by isolation and quarantine measures, may slow the spread of the disease. Also, we proposed a risk index, based on ranks of the estimated time for an infected individual to be identified in a specific city. The risk index was shown to be robust and consistent with the spreading patterns, independent of the mobility intensity parameter s .