A Spatio-temporal SEIRD Model of Covid-19 in Sweden

Abstract: The Covid-19 pandemic has affected many countries all over the world, and more than three million people have died from the virus. To understand the development of a pandemic it is common to use a simulation. This report investigates if it is possible to create a potential simulation of a spread of Covid-19 in Sweden, as well as in Stockholm, Gothenburg and Malmo, and how well it corresponds to the first 150 days of the pandemic. The report uses a compartmental model, SEIRD model which divides the population into: susceptible, exposed, infectious, recovered and deceased. This model is composed of a system of partial differential equations which are dependent on time and space, and these are solved with the finite element method. The simulation shows a possible spread of Covid-19 and the simulated number of deceased is compared to data from the Public Health Agency of Sweden. Comparisons shows that the model is not credible but agrees best with Sweden with a ranged- normalized RMSE value of 0.66 and worst with Malmö with a value of 2.53. For future studies the model could for instance use a population density based mesh that better corresponds to reality or initiate the virus in more places in Sweden.