[Manycouches] covid forecast of a (no)wave for Summer IETF in Philadelphia, USA

[Alexandre Petrescu <[e-mail]>]

Hi, SHMOO participants,

This email is about forecasting the covid wave (or no wave) peak for 
Philadelphia, during the Summer IETF of 2022.

Will the local covid wave be in a low, ascent, peak, plateau or descent 
during the meeting in Philadelphia?

Ideally, there would be no covid wave then and there, but one never 
knows.  Hence the need of a forecast.

The following two URLs illustrate forecasts that suggest that there 
might be no high wave of new cases in July 2022 in the Pennsylvania 
state of USA:

https://covid19.healthdata.org/united-states-of-america/pennsylvania?view=resource-use&tab=trend&resource=all_resources

https://www.covid19.uga.edu/nowcast.html

On another hand, the forecasthub provides many models for short term 
3-week forecast.  Their predictions vary from a strong descent to an 
exponential-like ascent of the wave until May 7th, 2022.  The URL is this:
https://viz.covid19forecasthub.org/

CDC is an authority in USA. It offers many covid forecasts, but lacks a 
display of a forecast of new cases at this time.  This is the URL:
https://covid.cdc.gov/covid-data-tracker/#forecasting_weeklycases

---

Personally, with a calculator at hand, I'd say the next wave peak of new 
cases would be on May 6th, 2022, with a correlation factor of 0.88 (1 if 
an exact fit, or highest confidence).  It is calculated using a 'Root' 
regression: ab^1/x of WP43S present on gitlab; it is fed with the list 
of the 5 earlier wave peak distances in Philadelphia district of the 
Pennsylvania state I could identify roughly from CDC, outbreak.info and 
worldometers: 244, 124, 155 and 117 (days between approximate peaks of 
April 11th, 2020, Dec. 11th, Apr. 14th, 2021, Sept. 16th and Jan. 11th, 
2022).

I did not use 'Ensemble' models, nor the R mathematical software 
package; also, I did not use specific covid-related epidemiology factors 
such as vaccination effect forecasts, masks mandates, lockdown measures 
or similar heuristics.

Alex

PS: another interesting wave forecast, but for UK, shows there might a
     peak of 'prevalence' (blue line, probably directly proportional to
     'new cases') on approx. Jan. 2nd, 2023.  It is at this URL:
     https://www.fil.ion.ucl.ac.uk/spm/covid-19/forecasting/

PPS: California forecasts as well, but not the case numbers.  This is
      the URL: https://calcat.covid19.ca.gov/cacovidmodels/