I remember well when in university I was first explained what a mock relationship is. In Denmark there was a strong correlation between the birth rate and the number of storks that had their nests there.
The explanation was that there are more storks in the countryside, where also more children are born than in the city. But of course storks do not bring babies.
A good lesson to realize that if certain things are numerically connected, it doesn’t have to be because there is a causal connection. So if you find a numerical connection, then the next task is to find out if there is a causal connection. By changing one variable, the other one should change as well.
So I briefly go through the steps that have been taken and the evidence that has been found. With the links to where that information comes from.
A. There is a numerical correlation between the specific humidity and the number of deaths caused by COVID-19. This was already found in this study in early March. In the meantime, a statistical analysis of 69 areas in Italy, the US and the Netherlands has also been carried out. In this research report.
you will find the numerical analysis of those regions, which shows, among other things, that if the specific air humidity was between 4 and 6 g/kg on average in the previous 3 weeks, a region reached 1 death per million inhabitants per day, twice as fast as if that value was higher or lower. So the correlation has been established. But that does not mean a causal connection.
- Several scientists are giving explanations, which could lead to this. Two possible explanations that are given are:
In this article, Prof. Evengalista (microbiologist and virologist) gives this possible explanation:
And there is also this explanation suggested by a group of scientists mentioned in this article.
In the U.S., the government commissioned research at the University of Utah to investigate both possibilities in the lab. I hope they come up with their results soon. Personally, I think it’s quite possible that both statements are valid, but time will tell.
In 2013 there has been large-scale research, based on data from 48 locations around the world, into the relationships between the development of a flu epidemic at those locations and the weather. For each location, several years of information was used. The result of this research was, that based on the specific air humidity, it could be predicted with more than 80% accuracy, that there would be a flu epidemic one month later. This is the graph that shows this relation well. It can be found in this article.
We see as well, that there is a clear relation between the specific humidity (the grey columns) and the start of the flu epidemic (the black dot). And that the limit of 6 g/kg is really visible.
But this is still no proof that there is a causal relationship. Other factors may play a role as well.
D. Subsequently, evidence has been provided in various ways, that a viral infection can occur via micro drops in the air (aerosols). The Japanese chairman of the association of virologists shows this in a video. Ejected micro droplets remain in the air for a long time and can infect those present. Here I report on it.
This would be an excellent explanation for the large-scale distribution (super spread events) at après-ski, carnival, large church gatherings, choir rehearsals, the distribution within care institutions (the air circulation system) and the distribution on cruise ships and navy ships. Incidentally, this is a way of distribution that is not (yet) recognized by the WHO and RIVM.
In that video, the Japanese professor indicates that ventilation is a good way to make these micro drops disappear. Based on the previous three findings, it is now very plausible that humidity also has an influence, on whether or not those drops float for a long time. Prof. Evangelista (see point B) says this literally.
The final proof came from animal experiments in which it was established that animals that released the influenza virus, at varying levels of humidity, infect other animals and the extent to which they do so. Professors of micro-biology and virology in Switzerland and at Yale University have recently written an article in which they report on past experiments on air humidity and the spread of the influenza virus. I am reporting on this here. (I now have intensive contact with them).
Several researchers have conducted experiments with mice, guinea pigs and ferrets. Some of the animals were infected with influenza. The uncontaminated animals were separated from the sick animals. So there was no direct contact. So the only possibility of infection was through the air (micro drops).
In the experiment the infection of the other animals was measured at different humidity levels. The researchers found that there was a clear correlation between the air humidity and the degree of spread of the virus. This picture shows their findings.
These experiments demonstrate conclusively that there is a causal relationship between the humidity and the extent to which one can be infected with a virus via aerosols.
So the correlation found, mentioned under A, is not a mock correlation but a causal relationship: the degree of specific humidity has a direct influence on the speed of spread of the virus.
Nevertheless, there may also be other factors that influence the speed of the virus spreading. Many possibilities have been put forward: the vitamin D level of the population, the UV content of the sun, pollen in the air and the fact that one is more indoors in winter.
I don’t explicitly exclude those, but there might just as well be a mock correlation with humidity (as I think is the case with pollen). And it would be nice if also with regard to those alternative explanations an equally hard proof would come, as through the evidence given by me.
Because any well-founded knowledge of how the speed of the virus spreading is slowed down is an important weapon in this battle.