Inferring COVID-19 spreading rates and potential change points for case number forecasts

Jonas Dehning; Johannes Zierenberg; Frank Paul Spitzner; Michael Wibral; Joao Pinheiro Neto; Michael Wilczek; Viola Priesemann

Detecting change points in the spreading rate -and quantifying the amount of change as quickly as possiblebecomes a central modeling challenge when short-term forecasts are required. To address this challenge, we assume an initial spreading rate λ 0 (the exponential growth phase, cf. Fig. 1 ) and up to three potential change points motivated by the German governmental interventions: The first change point (λ 0 → λ 1 ) is expected around March 9 (t 1 ) as a result from the official recommendations to cancel large events. A second change point (λ 1 → λ 2 ) is expected around March 16 (t 2 ), when schools and many stores were closed. A third change point (λ 2 → λ 3 ) is expected around March 23 (t 3 ), when all non-essential stores were closed, and a contact ban was enacted. We expect the behavioral changes introduced at these change points to unfold over a few days ∆t i , however, changes in duration can be partly compensated by changes in the onset time t i (see Fig. 2 E,F, scenarios). We chose priors for all parameters based on the information available to us up to March 28 (see Methods). On that date, the data were already informative about the first change point, and thereby helped to inform our forecast scenarios.