Before reading this please do a quick self-check. If you haven’t had a break in a while please take a break. This stuff can take a toll on the mind. There are some more photos to ease the mind here if you like.
Today’s update from the MSDH saw no significant change in the growth trajectory. The first plot below shows raw data from MSDH reports in orange. Where the orange plot ends is today. The basic projection model starts on the silver plot. The basic model error for today is 8%, which is to be expected for the limited availability of raw data.
An exponential model for estimating surge and timing is shown in the second chart. It is plotted in yellow and uses a scaling factor of 50 and a reduced transmission rate (compared to that of the raw data) based on data obtained from a recently published study of Chinese infector-infectee pairs. The study examines the basic reproduction number and serial interval (time between onset of symptoms of infector and onset of symptoms of infectee). The scaled projection model shown here is very loosely based on this data, as many simplifying assumptions must be made. The result is a conservative estimate of surge timing.
Please note that this is in no way a robust mathematical model. Local data to support such a model simply is not yet available. So this is a rough estimate. The value of it is that it shows how the largest surge of cases happens much faster than the projection based on the raw data, which is simply the number of positive tests.
The best interpretation of the exponential model (yellow plot) is not of actual numbers of infections but of its shape and timing. The shape indicates a fast surge. The timing indicates a surge much sooner than the raw data projection.
Positive COVID-19 Test Results (orange) with raw data projection model (silver)
Raw data + projection (orange/ silver) with exponential model (yellow) estimate of actual cases overlaid
The exponential model shown here makes no claims of accuracy of any kind and must not be reported as accurate. An accurate model of estimates is not possible at this time. Please direct any corrections, questions, or new and better data to info@jjj-engineering.com
Please stay inside and take care of yourselves. I’ve posted some more photos on the photography page.