{"references": ["Ferguson et al. (2020), \"Impact of Non-Pharmaceutical Interventions (NPIs) to Reduce COVID- 19 Mortality and Healthcare Demand,\"", "Hethcote, Herbert W. 2000. \"The Mathematics of Infectious Diseases.\" SIAM Review 42 (4): 599\u2013653. https://doi.org/10.1137/S0036144500371907.", "Lemke, Mathias, Liming Cai, Julius Reiss, Heinz Pitsch, and J\u00f6rn Sesterhenn. 2018. \"Adjoint-Based Sensitivity Analysis of Quantities of Interest of Complex Combustion Models.\" Combustion Theory and Modelling, July, 1\u201317. https: //doi.org/10.1080/13647830.2018.1495845.", "Systems Science, Johns Hopkins University Center for, and Engineering. 2020. \"Novel Coronavirus (Covid-19) Cases.\" https://github.com/CSSEGISandData/ COVID-19.", "Robert Koch Institut. 2020. \"COVID-19: Fallzahlen in Deutschland und weltweit\" https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus/ Fallzahlen.html."]}

Data assimilation is used to optimally fit a classical epidemiology model to the Johns Hopkins data of the Covid-19 pandemic. The optimisation is based on the confirmed cases and confirmed deaths. This is the only data available with reasonable accuracy. Infection and recovery rates can be infered from the model as well as the model parameters. The parameters can be linked with government actions or events like the end of the holiday season. Based on this numbers predictions for the future can be made and control targets specified. With other words: We look for a solution to a given model which fits the given data in an optimal sense. Having that solution, we have all parameters.

Final Draft. CSV Files of prognosis data and updates to the figures will be published here: https://www.zenodo.org/record/3733256 v.1.2 Errors corrected. v1.2.1: More typos (Thx to Phillip Krah!)