Initial time in model should not be zero but the infected person first presence time
This has some importance since the introduction of hourly dependence of temperature, so in the calculator @mrognlie @pelson . Otherwise the model will start at midnight with probably quite some slow-down for the computation of useless hours from midnight till 8 am. This is in https://gitlab.cern.ch/cara/cara/-/blob/master/cara/models.py#L485