Solution of SIR Infection Equation Using Data Assimilation

Abstract

The new coronavirus infection (COVID-19) is rampant. The most troublesome part of this infection is the time between infection and onset and the infectiveness for several days even in the not-onset state. Therefore, a considerable number of infected persons with infectivity are left unchecked. Therefore, even if the infection status is simulated by the SIR equation or the like, the true values of the infection parameters and the true number of infected persons cannot be grasped. However, it is possible to observe the infection status, and the daily number of infected people and the cumulative number of infected people are announced. These numbers are not true values, but they reflect true values. It is impossible to grasp the true value only by the SIR equation, but it will be possible to estimate the true value by combining it with the observation equation. In short, the data assimilation framework is considered to be effective. We report this effectiveness because we were able to confirm this effectiveness from the numerical results.