For the first time a direct correlation (association) between relative humidity and number of Covid-19 cases (waves) has been proposed using graphs with data from London (Note, the graphs are ‘live’- updated regularly and independently verifiable)
Although scientists have long known relative humidity influenced transmission of the virus that causes Covid-19, SARS-CoV-2. a direct easy to reproduce graphical representation between humidity and Covid cases has not been shown until now.
Background:
The discovery was promoted by preliminary findings that suggested low relative humidity values being observed prior to a number of surges happening around the world in Delhi, South Africa, U.K. and other countries.
As the outdoor relative humidity goes up, cases are observed to go up and when they are low cases are also lower.
Note however relative humidity in and outdoors is not usually the same especially in the winter when conditions outside are not the same as indoors.
(This is explained in more detail below under Relative Humidity In and Outdoors.)
The lowest values in relative humidity appear to have a relationship to Covid-19 infection rates and follow the same wave pattern as the rate of infection. Normal daily variations between the minimum and maximum relative humidity readings appear to have little influence, with only the minimum values in any given period being relevant to this correlation.
Minimum humidity data from London since February 2020 to present was used to select the lowest relative humidity values in any given two-week period. Up to three of the lowest values were chosen (choosing one to three of the lowest minimum humidity values selected within any 14-day consecutive period is adequate for this exercise) All relatively higher humidity values were ignored. As will be seen later in several scientific publications and research papers, periods when relative humidity are low represent the worst case scenario for transmission and infection of SARS-CoV-2.
A graph of the results was plotted against case numbers as recorded by the U.K. Government website since February 2020 to 21 August 2021 (now extended to date). The results are shown on the following graph.
Owing to the relative success of this model over the past 18 months, the author will continue to update the graph with new data from U.K. Gov and London Weather Centres daily. Please check the MS Excel link under “data and references”.
Correlation with Covid Hospitalisations in LONDON
The same correlation was developed for Covid hospitalisations in London using the same minimum humidity values. The London Covid hospitalisations data was sourced from UK Gov website using the following link:
https://coronavirus.data.gov.uk/details/healthcare?areaType=nhsRegion&areaName=London
Note UK government has provided two separate sets of Covid hospitalisation numbers:
The first set of data is for new admissions and can be found in the following link: https://coronavirus.data.gov.uk/details/healthcare?areaType=nhsRegion&areaName=London
The second set of data is for hospitalisations available in the following link.
https://coronavirus.data.gov.uk/details/download
All sets of data have been utilised and the graphs produced can be found in the MS links further below under “data and references.”
Interestingly protections such as vaccination, lock downs, mask mandates etc do not appear to have affected the correlation which covers the entire pandemic period to date?
One of the unique advantages of this correlation is anyone can independently reproduce their own copy of the graph easily and quickly! A simple step by step guide is provided further below on how this can be achieved. Data used to produce the graphs above can be found at the end of this article. As will be seen later the relationship is useful in predicting Covid 19 infection waves. It may also potentially provide alternative approaches to defeating the SARS-CoV-2 viruses.
Relative humidity in and outdoors:
Technically, in the winter high relative humidity levels outdoors (saturated air) can result in very dry conditions (unsaturated air) INDOORS. This is due to temperature differences between outside and inside. Although relative humidity outdoors maybe high, the low temperature means absolute humidity (amount of water in vapor form per cubic metre of air) is actually very low. In other wards low temperature simply means lower water content (in vapour form) but higher water saturation or higher relative humidity. When this air gets indoors, absolute humidity remains very low but suddenly the higher temperature causes relative humidity to drop significantly as the air becomes able to absorb a much higher amount of water due to being unsaturated. (This is why laundry dries rapidly indoors during the winter compared to outdoors). People tend to spend more time indoors in the winter. This indoor “dry air” or (low relative humidity) maybe the reason for higher Covid infection rates as explained later. This could be one reason why ventilation might reduce infection rates. Ventilation will cause indoor and outdoor relative humidity to be nearly the same.
The same low relative humidity conditions can arise in mid-summer in the months of June and July in the UK and many other countries (examples given later in this article), in summer it is possible for the outdoor temperature to rise leading to relative humidity (and absolute humidity also) to drop down to 20% and lower. This maybe the reason for surges around this time, which can be exacerbated by extensive use of Air conditioners which are known to be excellent dehumidifiers. Use of the air conditioners in a car illustrates the mechanism and speed of change between saturated and unsaturated air. Turning it on in a saturated environment with condensed water on the windscreen results in the evaporation of the surface water within a few seconds. ( Essentially air conditioners work by venting very cold dry air, this cold air naturally has very low absolute humidity due to it’s low air temperature. This then enters the cabin environment which has a relatively higher temperature, and converts it from a saturated to an unsaturated environment that ‘sucks’ up any moisture on surfaces in a process that is fairly rapid)
As shown below research by the US Dept. of Homeland Security and other reputable institutions show that stability of the SARS-CoV-2 virus is significantly affected by variation of relative humidity.
1.
2.
3.
4.
Additional relative humidity research and studies
Below are links to additional research and studies many done by leading scientists and researchers on the relationship between the SARS-CoV-2 virus and relative humidity. Note most of them were produced during the early part of the Covid pandemic but once Covid vaccines were introduced interest in this area seems to have ceased? Note that a number of other respiratory viruses besides Sars-CoV-2 appear to be susceptible to variation in relative humidity.
5.
How Seasonal Variation in Indoor Humidity Affects COVID-19 Transmission | BNL Newsroom
6.
7.
8.
https://uia.brage.unit.no/uia-xmlui/bitstream/handle/11250/2990521/Article.pdf?sequence=4
9.
10.
Using Machine Learning to Estimate COVID-19’s Seasonal Cycle (lbl.gov)
11.
12.
13.
14.
15.
16.
17.
The Role of Dry Winter Air in Spreading COVID-19 | University Hospitals (uhhospitals.org)
18.
19.
20.
Seasonality of Respiratory Viral Infections | Annual Review of Virology (annualreviews.org)
21.
22.
https://www.nature.com/articles/s41467-020-16670-2.pdf
23.
24.
25.
26.
Frontiers | Effects of Environmental Factors on Severity and Mortality of COVID-19 (frontiersin.org)
27.
https://www.nature.com/articles/s41598-021-91798-9
28.
https://royalsocietypublishing.org/doi/10.1098/rsif.2021.0865
29.
30.
31.
32.
33.
34.
How this correlation may help explain the Covid wave patterns:
As observed in the links above, research shows humidity affects SARS-CoV-2 ability to:
2. Transmit (3, 4, 5, 6, 7, 9, 11, 13, 14, 16, 17, 27, 29, 33)
3. Infect (3, 8, 15, 18, 26, 28, 30,)
{4. Note: Variation of humidity can also have an effect on influenza
For example, according to US Dept of Homeland Security’s calculator, increasing relative humidity from 20% to 60%, reduces virus stability by at least 40%.
The findings generally suggest dry air or low relative humidity result in higher concentrations of the airborne virus indoors for longer periods. (It maybe that the virus is present on surfaces mainly when conditions are humid and more likely to be airborne during drier conditions – See appendix at the end of this write up).
As will be demonstrated later, the correlation graph was developed based on this premise i.e., that low relative humidity generally increases the rate at which Covid infection occurs.
The implication is infection is more likely to occur during times when relative humidity is low, which could literally be a few minutes or hours during mid-summer (May to July) when weather charts show relative humidity dropping to as low as 20% or for longer in the winter when minimum relative humidity in indoor spaces drops to its lowest levels (as happens seasonally in December in many parts of Europe).
When minimum relative humidity is continuously above a threshold, the concentration of airborne SARS-CoV-2 viruses drops, the ability of the virus to survive on surfaces reduces, transmission reduces and Covid cases plummet leading to low prevalence.
Outdoors it is blown away, dispersed in the air which reduces its concentration or presence. This also explains why transmission mostly happens indoors because the virus becomes trapped circulating extensively when conditions are favourable.
Altogether it may also show why current physical controls do not appear to make much difference to the wave patterns because when the conditions are conducive for the virus (low relative humidity/dry air) the virus simply spreads indoors easily and rapidly, survives longer breaching many of the current controls, causing cases to increase and vice-versa, resulting in a ‘seasonal’ wave pattern.
Natural variations in minimum relative humidity are generally governed by seasonal changes as shown by weather charts. Examples for 2019, 2021 and 2022 are shown illustratively below from an amateur London weather station. (Note that these are outdoor measurements which reflect similar relative humidity levels indoors only in summer but not necessarily in winter. This is because in the winter indoor temperatures could be much higher than outdoors when heaters are running. This higher indoor temperature will lead to significantly lower relative humidity indoors. For example, it is observed laundry dries rapidly indoors in mid-winter because of this very low relative humidity)
Although variation in minimum humidity is influenced by seasonality, it is still possible to artificially control humidity in an indoor setting. This does not require an energy source in the winter. It is common knowledge water from laundry evaporates rapidly indoors in the winter implying any material with a large surface area containing water will humidify a room even without an external energy source.
Significance of this characteristic with respect to other coronaviruses:
As this phenomenon is related to the physical structure of the virus (ie do their spikes increase capacity to hold water molecules?), it will likely apply to other coronaviruses:
1. https://www.nature.com/articles/s41370-022-00472-3.
2. Dry air is known to enable viral transmission of influenza.
Above the rough representation of the physical structure of coronaviruses surrounded by spikes may make them have an affinity for water molecules. This has the potential to change their combined densities depending on the environment ie when its dry or humid. A dry environment will result in potentially lower densities and a humid environment higher densities which likely affect the viruses ability to be airborne.
For example, although Covid cases appear to have been relatively low for the season in December 2022 for London, the number of flue cases nevertheless was high overwhelming several hospitals:
NHS England » Flu pressure rises with hospital cases up seven fold in a month
Data and References:
The following Excel spreadsheet is updated regularly with new data from U.K GoV and London Weather centres and will be occasionally cross-checked with Zoe Covid data or any other relevant data. For the latest updated graph please check the excel spreadsheet on the link below. (Note, this depends on when and how frequently UK Gov updates its data which at present is 4.00pm U.K time on Thursdays). As already stated, the spreadsheet includes raw data for the hospitalisation graph above in addition to the Covid case numbers easily verifiable and independently reproduced.
NOTE: UKHSA NO LONGER PRODUCES DAILY COVID DATA SPECIFIC TO LONDON AS OF 14/DEC/2023 COVID DATA UPDATES TO THE FOLLOWING LINKS ARE THEREFORE NO LONGER POSSIBLE FROM THAT DATE. HOWEVER RELATIVE HUMIDITY UPDATES FROM THE LONDON WEATHER STATIONS WILL CONTINUE.
https://coronavirus.data.gov.uk/details/cases?areaType=region&areaName=London
Links to MS Excel graphs:
Correlation (i) produced by London Heathrow weather Station (https://www.wunderground.com/history/monthly/gb/hounslow/EGLL/date/2020-2);
Correlation (ii) produced using weather charts from a London amateur weather station: (http://nw3weather.co.uk/wxdataday.php?vartype=hmin)
I have used the new dashboard data for England to plot a graph against minimum relative humidity values from weather charts produced from London Heathrow airport for my most recent correlation graph which can be found here:
England association Graph with minimum humidity values.xlsx
The UKHSA data can be found in the following link.
https://ukhsa-dashboard.data.gov.uk/topics/covid-19#cases
UKHSA surveillance for Influenza cases, admissions and hospitalisations is produced weekly. The latest data can be found here. It shows the anticipated surges in Covid and Influenza cases for December 2023 during the 51/52 week update. See graphs below.
Source for this graphs is in the following link: https://assets.publishing.service.gov.uk/media/65843c4123b70a000d234db8/Weekly-flu-and-COVID-19-surveillance-report-week-51.pdf
(See https://link.medium.com/naV6rT1KUkb for details on Scotland)
On another note, Zoe Covid study is an organisation independently collating and analysing U.K. Covid data. The following is a U.K. wide graph of Covid case numbers over several months. Note how it appears remarkably similar to the minimum humidity (green) graph above despite London data only used to produce it.
GERMANY data correlation graphs
The graph above is the same minimum humidity verses case numbers for Germany. As with Scotland one weather station was used to obtain humidity values ie Berlin Airport, nevertheless again a correlation is observable for the period since Feb 2020 except March to April 2021 where there is a surge with no corresponding rise in minimum humidity values. The following excel spreadsheet contains the raw data used to develop the graph. The weather data can be found in the following link: https://www.wunderground.com/history/monthly/de/sch%C3%B6nefeld/EDDB/date/2020-6
India Covid (wave) surge in cases March 2023
A Covid surge in new cases occurred in India during the months from March to April 2023, peaking in the middle of April prior to cases starting to decline in May 2023. The cause was broadly believed by experts and scientists to be because of the Omicron variant XBB.1.16 also referred to as Arcturus. This variant was discovered in at least 30 countries with only India and to a lesser extent Singapore experiencing a surge of Covid cases as a result. So far, no explanation has been given as to why no other countries experienced a surge (wave) in cases as a result of the presence of XBB.1.1.16 as occurred with the Delta and Omicron variants in previous years. Also, no explanation as to why the surge peaked when it did mid-April prior to cases dropping in May 2023. Several articles were published concerning this variant at the time including:
https://www.independent.co.uk/news/health/arcturus-new-covid-variant-india-uk-b2319005.html
As has been done before it is possible to demonstrate how low humidity (dry air) concept may have an association with the surge in Covid cases more importantly a verifiable prediction of precisely when the surge peaked and why no immediate surges (waves) occurred in countries where this variant was present. It will be shown that the highly favourable conditions for the virus to transmit, survive and infect according to several studies (low humidity/ dry air) was in existence prior to this surge in India with humidity values dropping below 10% over a number of days early February 2023 followed by below 22% in March which represented very dry air conditions demonstrated through several studies to be highly favourable conditions for SARS-CoV-2 infections to increase leading to surges. Links to some of these studies have already been provided above.
The following are tweets made during the course of this surge which among others predicted precisely when this surge would peak and may even satisfy elements of the Bradford Hill criteria.
- Proof of very dry air conditions existing in Maharashtra India one of the epicentres prior to this surge early February 2023 derived from weather records:
https://twitter.com/MikeKeelfree/status/1645502923252842496?s=20
2. The accurate prediction of when this surge peaked presented within a thread introduced by:
https://twitter.com/vipintukur/status/1646351684938129409?s=20
3. Other related tweets:
4. Easy access to India daily regional Covid data:
https://twitter.com/outbreak_india
https://twitter.com/MikeKeelfree/status/1651988753911541762?s=20
London Covid predictions:
Some of the predictions made using the association between Covid infections and variation in seasonal humidity can be found in the following link:
https://twitter.com/MikeKeelfree/status/1664950441430401024?s=20
South Africa
As shown by the graph, South Africa has had surges in Covid cases in May every single year since the pandemic began. All these surges (May) peaked in July like clockwork. It is currently experiencing another surge this month (May 2022), will this peak in July like previous years? Something worth monitoring…
The above graph represents ‘our world in data’ covid cases data for South Africa in June 2022 and one below has been updated on 18/08/22. The raw data can be checked and verified by ‘our world in data’ here:
covid-19-data/public/data at master · owid/covid-19-data · GitHub .
The graphs were prepared using the same MS excel sheets above and can be found in one of the excel sheets.
Interpretation of the graphs is open to anyone.
It appears a report by the World Health Organisation (WHO) is reflected in the graphs, it suggested in April 2022 that a running decline in Covid cases in Africa existed:
Africa witnesses longest-running decline in COVID-19 cases | WHO | Regional Office for Africa
Vaccination rates for South Africa as of 18/08/2022 - 32.5% of the population is fully vaccinated according to ‘our world in data’:
Coronavirus (COVID-19) Vaccinations — Our World in Data
Mask wearing is no longer required as of June 2022:
South Africa repeals COVID rules on mask-wearing, gatherings, entry — CNBC Africa
USA and NYC Twitter Covid predictions:
US data beginning with NYC. Will be presented in a new medium article at a future date.
The following graph was produced using a correlation with Dew Points which is a good measure for vapour content. Please find a work in progress NYC graph below…
The following predictions were made for New York City on Twitter and all came to pass …
Prediction 1 (May 11 2022):
Conditions are favouring SARS-CoV-2 . (See thread for this prediction it states: Suggests probability of infection in New York City is currently quite high. If I were in NY, I would take my usual precautions, essentially ventilation and use of some form of cold humidifier when indoors) (Note there was no way of predicting this using Covid case data as cases were low at this point with no sign of any surge)
Soon after this prediction the virus Alert level was raised from Medium to High on 17th May 2022 by NYC.GoV:
Prediction 2 (May 31–2022):
Conditions have changed and are now against SARS-CoV-2 (Note the reference date of change on the graph was 18–05–2022).
A few days later cases were shown to level off, CDC reduced the alert level to Medium on 04/June/2022. NYC was still maintaining High alert level but was predicted to downgrade by the end of June or early July 2022:
CDC downgrades alert level on (June 4 2022):
- New York City downgrades alert level to medium
NYC data and Charts
NYC archive weather Dew points data were obtained from the following website:
https://www.wunderground.com/history/monthly/us/ny/new-york-city/KLGA/date/2022-4
Source for NYC Covid hospitalisations data were obtained from the following website:
Follow instructions to download the CSV zipped files under “code”. Open the “coronavirus-data-master” folder and select trends followed by hosp-by-day file.
The following link contains the MS spreadsheets and correlation graphs including all formulas to produce/reproduce the correlation:
The London graph shown can easily and quickly be reproduced by anyone using the following simple steps! (Note the same procedure is used to produce all other correlation graphs from different countries)
To reproduce the graph:
- Go to the U.K Gov website using the following link:
a) For new daily Covid cases:
https://coronavirus.data.gov.uk/details/cases?areaType=region&areaName=London
b) If you want to use Covid hospitalisations in London, use data from the following link. (Note: The procedure for developing the graphs is exactly the same in both instances for new Covid cases and hospitalisations):
https://coronavirus.data.gov.uk/details/healthcare?areaType=nhsRegion&areaName=London
https://coronavirus.data.gov.uk/details/download
Note: As explained earlier, UK Government provided two different sets of daily London hospitalisation data on their website, one set for new admissions and another hospitalised numbers. The can be found in the links below and can both be used for this excercise.
2. Download the U.K cases data found below the graph and choose CSV.
3. Use this as your Excel Spreadsheet template. Input data from the NW minimum humidity weather link (Find the link at the bottom of this page and click to get the raw data set, note that this data set is used for illustrative purposes only as it presents all minimum humidity data for one year on one page otherwise use data from any other weather stations such as Hounslow, United Kingdom Weather History | Weather Underground (wunderground.com)).
Only the lowest values should be selected and fed into the spreadsheet in their respective days. Any high values can be ignored. For example, if the values you see are: 92, 31,73, 33, 20, 67, 46, 59, 90, 46, 88, 42 pick only 31, 20, 59, 33, 46 and 42 to place on your spreadsheet. A simple example is shown below for months January and February 2022. (Note: Selection of only two to four of the lowest humidity values in any consecutive circa 14-day period is adequate for this exercise.)
*These values are referred to as the influential minimum humidity values. (As already explained earlier, this model relies on days when minimum humidity is lowest as these values represent the worst-case scenarios for Covid infection rates based on findings (see links to this research above) that show survivability, transmission and infection of the SARS-CoV viruses are highest when humidity is lowest and vice-versa. More details on influential humidity values and how to select them can be found here: https://medium.com/for-the-first-time-the-relationship-between/living-with-covid-f0aed00e0ce5).
Input influential values in their respective dates using one column of the spreadsheet. Work your way month by month. Call this column h.
4. Apply exponential empirical function ( h*h) where h are the values you’ve fed in h. Apply this function in next column for all values in h. Multiply h*h by 2.7 on another column as “amplification” you can use any other value of your choice as amplification so that your results are visible on the y — axis of the graph.
5. Produce the graph for the Y- axis – New case numbers (column E) and (h*h)*2.7 column that you have just produced against Days on the X-axis (Column D).
6. The case numbers as a line graph and humidity as bar graphs.
7. All done!
Limitations of using weather stations from different locations around the world:
For accurate comparison between relative humidity and Covid case numbers to the raw data or input data needs to be relatively accurate. Unfortunately this is not always the case as for example some weather stations do not report readings regularly, with intervals ranging from less than 5 mins to over 3 or 4 hours. The ideal readings to obtain minimum humidity values need to be collated in less than 30 minute periods to avoid missing the critical low values that are thought to trigger surges. https://www.timeanddate.com/weather/forecast-accuracy-time.html
APPENDIX:
INTRODUCTION
1. CORONAVIRUSES and WATER MOLCULES
SARS-CoV-2 is a coronavirus. Note that several other coronaviruses exist besides SARS-CoV-2. Many influenza or respiratory tract disease causing viruses for example are coronaviruses. They are named coronaviruses because of the spikes surrounding them that appear like a crown. SARS-CoV-2 individual size ranges from 50 to 140 nm.
Water molecules physically exist in different forms such as liquid, water vapour (gas), steam and ice. In this discussion will be more interested in water in its natural gaseous state.
PHYSICAL PROPERTIES
Due to their relatively large surface areas coronaviruses appear to provide suitable areas for absorption and temporary retention/accumulation of water molecules. The picture of the towel is there to try to illustrate this point. The typical size of a water molecule is 0.27nm. The graphical representations below illustrate the difference in size between a water molecule and SARS-CoV-2 viruses.
SIZES COMPARISON 1:
SIZES COMPARISON 2
DRY AIR CONDITIONS:
HIGH ABSOLUTE HUMIDITY CONDITIONS:
Note: Commercial dust suppression equipment depends on similar principles.
WINTER CONDITIONS:
MID SUMMER CONDITIONS:
A brief background:
I started doing research and looking at different data sets regarding SARS-CoV viruses and Covid during the U.K. lockdown of March 2020. I was having restless sleepless nights in any case so what better way to pass the time than trying to understand what had brought the entire world to this point? My gut feeling based on my faith as a Christian and the history of pandemics and epidemics was there had to be a simple vulnerability that could easily be exploited on the Sars-Cov-2 virus. Why? If you look back at epidemic/pandemic history starting with the black deaths of 1346 to 1353, that resulted in casualties of over 75 million people. It was eventually found to be caused by the bite of an infected flea and spread by pests, predominantly rats. Which means it could have been easily and was eventually effectively dealt with simply by pest control and sanitation measures! Cholera which also had casualties in the several thousand during pandemics that occurred between 1817 to 1923 was eventually found to be controllable by simple sanitation measures and also effectively treated with infusion of saline solutions and finally Typhoid fever was discovered to be preventable and controlled by simply good sanitation and hygiene.
With that in mind I commenced my research spending hours on end beginning from scratch. I completely had absolutely zilch, no idea what enveloped viruses where, or what Covid was or even what I was doing? I just looked at everything factual, reliable and verifiable that I could find. Early on I noticed that there was mention of the virus being impacted by variation in temperature. Also there seemed to be more likelihood of surges in winters as compared to summers and hot tropical countries had seemed to fair much better than expected compared to temperate countries. There was also data by none other than the US Dept. of Homeland Security website showing the half-life of the Sars-Cov -2 viruses were affected by variation in temperature and relative humidity. So, I therefore put more emphasis and attention in that area looking for credible material and data sets that I might use to gain more knowledge on this. I would spend several hours researching even taking time off work for the purpose. Eventually in August 2021 while on holiday in the Peak District (England) I achieved my “Eureka” moment. I was as usual looking at raw data and producing various graphs trying to discover if a correlation existed between temperature and Covid case numbers and I included humidity data as it might give me some additional insight as I had noticed surges had often occurred following low humidity values, as usual I couldn’t find any anything, a pattern, anything useful – nothing. Suddenly my eye caught something, the minimum humidity graph which I was ignoring for the most part seemed to have some similarity to the Covid case number graphs! And with just a few tweaks and basic changes in my formulas it was suddenly there staring at me. I looked at it in disbelief and it seemed to stare back at me. A clear direct correlation between minimum humidity values and Covid case numbers for the London area.
Gratias ago tibi Domine propter sapientiam tuam.
Omnis sapientia a Deo est!
© Copyright
All Rights Reserved
