Covid-19: How Seasonal is it?

In my previous post I mused that: “…if the seasonal effect is not extreme, the UK might achieve sustained herd immunity to the Delta variant before Christmas in 3 months time.” It’s long been known that respiratory viruses, such as colds and flu, are highly seasonal. But surely never before has the spread of a virus been tracked so accurately as Covid. Consider this figure, a screengrab from the UK government’s dashboard:

What strikes me about this figure are the declines in cases simultaneously across all age groups in late July and early September. Something must have affected the whole population at the same time. Was it a lockdown? No. In fact restrictions were loosened in late July and the decline in case rates made fools of many epidemiologists who had warned of a further increase.

The most parsimonious explanation is that the late July heatwave had a decisive effect. The data now shows a further sharp decline in cases in early September, also coinciding with a heatwave.

Going back earlier, dips in rates at the start of March and April also followed unseasonally warm spells.

It would be good to know precisely what it is about cooler weather that favours the virus. Does it simply survive longer in the environment? Is the effect of temperature (and/or other weather parameters, such as humidity) on human physiology the cause? Are changes in our behaviour also important? The point is that accurate answers to these questions would give us some idea of the shape of the curve of infectiousness, that is, of the effect on R, of changes in the weather. Is there, for example, a temperature below which further declines no longer favour the virus?

The point is that as we head into winter we’re seeing a battle between conditions that seem to increasingly favour the virus (as well as continuing evolution of the virus with Delta strain AY.4.2 apparently even more transmissible than vanilla Delta) and increasing immunity of the population, due to continuing vaccination and the high level of infection.

Luckily this is a blog and not a scientific journal, so I’m allowed a guess. And that is that the current increase in case rates will be relatively short-lived – weeks rather than months – because of the rapid increases in immunity to infection of at least 1% of the population a week that the UK is currently seeing. 1% of the total population is, of course, much (maybe 3x) more than 1% of the remaining susceptible population. And I suspect the virus will see diminishing returns from the weather cooling further. But I could be wrong…