WSEAS Transactions on Computer Research
Print ISSN: 1991-8755, E-ISSN: 2415-1521
Volume 8, 2020
Numerical Simulations of a Modified SIR Model Fitting Statistical Datafor COVID19
Author:
Abstract: We consider a system of ordinary differential equations obtained by modifying the classical SIR model in epidemiology in order to account for the particular features of COVID19 and the structure of the available statistical data. Its main feature is that the infectious state is being split in two different stages. In the first one, which lasts a few days after being infected, the individuals are considered to be contagious and able to spread further the disease. After this, the individuals are considered to be isolated and this second stage lasts until either recovery or death is reported. The parameters of the model are fitted for several countries (Germany, Italy, Spain, Russia, USA, Romania) such that the solution matches the known number of new cases, active cases, recoveries and deaths. The values of these parameters give insight regarding the evolution of the pandemy and can reveal different policies and approaches in reporting the official data. For example one of them can indicate that in certain countries a substantial amount of cases were reported only postmortem. The variation across several countries of another parameter, which models the average convalescence time (the duration of the second stage of the infectious state), points to the fact that the recoveries are reported at different rates, in some cases with significant delays. Since it can be assumed that this is only a matter of reporting, we also perform additional simulations for these countries by taking the average convalescence time the value of Germany, which is the smallest within the whole range. The conclusion is that under this assumption, the evolution of the active cases for example in Italy and Spain, is not significantly different to that in Germany, the comparison being based on the fact that these countries showed a similar number of cases within the considered period.
Search Articles
Keywords: ordinary differential equations, SIR model, parameter fitting, numerical simulations, COVID19
Pages: 115-125
DOI: 10.37394/232018.2020.8.15