CORONA VIRUS (COVID-19): The Effect Of The Ambient Temperature

Table 1- Chromatic ranking for the different analyzed factors and for the COVID-19 cases

COVID-19 is a new virus. At moment, no one really knows how to treat it and we hope that a medical solution will be find soon. Meanwhile, the analysis of the published data can help to understand if there are factors that accelerate or reduce the contagion´s spread.

In the first days of the pandemic, I summarized the situation in the Italian regions and different worldwide nations ( I used a chromatic ranking to highlight the differences between the selected areas and to indicate the incidence of each factor: VERY BAD (Red), BAD (Orange), MODERATE (Yellow), GOOD (Green), VERY GOOD (Blue). In this article, I continue to use this approach: for each factor, each category is defined by a specific range of values. I also decided to apply the same chromatic approach for the COVID-19 data published by WHO and for the calculated percentage versus the population (see Table 1).

The following analysis focuses on the main factors that facilitate the contagion´s spread: population, population density and number of metropolitan cities (cities with more than 1 millions and 100000 inhabitants). The latter is a new factor, introduced to be more specific for countries (like USA, China, Russian Federation), where the population density results low due to the extension of the territories.

The incidence of these factors depends on the application and respect of quarantine and social distancing ( All the other factors discussed in former articles (% over 65, % smokers, % overweight, % obesity, health expenditure per capita, etc.) are still valid, but they have relevance for the fatality rate, which is not part of today´s analysis. Thus, I excluded them from the attached tables.

Moreover, following some discussions and generic analysis read in the first days of this pandemic, I spent time to evaluate the influence of an external factor: the ambient temperature. The analysis of the average value for the overall nation would not be the correct approach since, from north to south, from east to west, the situation could be dramatically different. Thus, in order to be more precise, I investigated the variation of temperatures for those provinces or cities that show high numbers of COVID-19 in each nations. Of course, I also introduced a chromatic ranking for this specific factor (see dedicated column in Table1), evaluated for the first three months of 2020.

Once completed this introduction, let´s see what I found and summarized in the following tables. I added my general conclusions at the end of the article.


Spain and Argentina

Table 2 – Summary table for Spain and Argentina. Top part of the table shows national data. Bottom part of the table shows local data for Madrid and Buenos Aires (COVID-19 Cases refer to the Province or Region of the Capital city)

The two countries have similar population, similar distribution in terms of metropolitan cities, relatively low population density according to my ranking (it is higher for Spain) and similar cultural, traditional and ethnic aspects. Therefore, it would be expected a similar evolution of COVID-19 cases.

This is not the case, with Spain well above 100000 cases and Argentina just above 1000 cases.

The first observation is that the first cases were confirmed in different periods, respectively on January 31st in Spain and on March 3rd in Argentina. Therefore, there is about 1 month of delay among the two nations. However, after reaching the first 100 cases, it took only one week for Spain to rise at 1000 cases (March 10th); it took 30 days for Argentina to rise at 1000 cases (March 31st).

What is the factor that is slowing down the rate of contagion in Argentina? Looking at the two areas (Madrid and Buenos Aires) which account for the highest cases in the respective nations, it is not the population or the population density: Buenos Aires would be clearly the underdog. Instead, it is interesting the comparison of the ambient temperatures with the Argentinian capital much warmer than the Spanish one in these months. Without any scientific proof and more investigation on the current behavior of inhabitants in Spain and Argentina, it seems that the ambient temperature plays an important role to slow down the contagion. Note: to slow down, not to stop!


Germany and Republic of Iran

Table 3 – Summary table for Germany and Republic of Iran. Top part of the table shows national data. Bottom part of the table shows local data for München and Tehran (COVID-19 Cases refer to the Province or Region of the Capital city)

The two countries have similar population, similar distribution in terms of metropolitan cities, and they were two of them I have been tracking accurately since the begin of this pandemic.

Germany has higher population density. Thus, according to this parameter, Germany should show a higher number of cases. This is effectively what the current status shows in April. However, this was not the status for long time.

In particular, Germany only overpassed Iran about two weeks ago (March 23rd). The reason is that the daily increase of cases in Iran has flattened in the last weeks (about 2500 daily), while those in Germany have doubled and still increase between 5000 and 6000 new cases daily (note that both countries are adopting social distancing and standard protocols to reduce the contagion). While I was not surprised about the situation in Germany (similar to other increase in Europe), I asked myself what in Iran was changing. I looked again to the ambient temperatures for two relevant cities (Münich and Tehran) and I was surprised to observed a dramatic change of temperatures in Iran: March is warmer than February by about 6°C (with peaks at 22°C). It is also worth to mention that, in March, in some days the temperature in Münich was well below 0°C degrees (-7°C) while this never happened in Tehran.

Moreover, I decided to compare the internal situation in Germany between Münich/Bayern and Frankfurt am Main/Hessen. Both cities/areas are highly populated and with busy airport and transportation systems. Well, the COVID-19 cases are respectively 20962 and 4097. At the same time, Münich was 4°C colder than Frankfurt am Main in the last month. Therefore, even an internal comparison among the German regions indicates the important role of the ambient temperature.

Like for the first comparison, the conclusion is again that higher ambient temperature might help to slow down the contagion´s spread.


Norway and Singapore

Table 4 – Summary table for Norway and Singapore. Top part of the table shows national data. Bottom part of the table shows local data for Oslo and Singapore (COVID-19 Cases refer to the Province or Region of the Capital city)

Singapore is one of the countries where the first COVID-19 cases are documented: 23rd January 2020. Moreover, it has one of the highest population densities and one of the busiest airports in the World. Indeed, it is one of those countries with very accurate controls (I can confirm since I travelled there in March). However, I could not explain myself why the rise of cases was luckily so slow: basically, more than 2 months to reach 1000 cases, while these values are the daily new cases in many European countries!

Thus, I decided to compare Singapore with the Norway, one of the European countries with a similar population (about 5 Million of inhabitants) and indicated by other studies as “in a good situation”. Well, first of all, I am not so much in agreement with this statement. Namely, the absolute values in Norway are still relatively low (4935). No doubt about that. But we have to consider that the country has low population density (17 inhabitants/km2) and no metropolitan cities. As consequence, I would expect much lower values of cOVID-19 cases. Instead, the ratio of the cases versus the population is one of the highest in Europe and it is almost 5 times higher than Singapore. We do not have to forget that the first case in Norway was documented on 26th February 2020. Namely, the current status in Norway is the result of the progression in only 35 days.

Therefore, I looked at the ambient temperature in the last three months. No surprises. The faster rate of cases of Norway could be explained by the colder temperatures; conversely the slower rate of cases of Singapore could be explained by the warmer temperatures.


Japan and Russian Federation

Table 5 – Summary table for Japan and Russian Federation. Top part of the table shows national data. Bottom part of the table shows local data for Tokyo and Moskow (COVID-19 Cases refer to the Province or Region of the Capital city)

Since the beginning of COVID-19 crisis, Japan has always showed a very slow progression of cases, despite the fact it was one of the first to document a COVID-19 case (early January 2020). Moreover, it is population, population density and cities sizes should be unfavorable to contain the spread. In particular, the metropolitan city of Tokyo is the most populated area of the world. On the contrary, the traditional discipline and the attitude of Japanese population to wear protective masks is a positive important factor to contain the spread. Moreover, the country is an island and this aspect also facilitates to contain external negative influences.

The Russian Federation is very similar to Japan for the overall population and for the city’s sizes. Some protection´s rules are also firmly applied in Russia. However, it differs significantly for the population density, that is much lower (9 inhabitants/km2) due to the large surface areas of its territory. The latter factor should bring less cases in Russia than in Japan, which is not the case. So, how can we explain the current differences between Russian Federation and Japan, considering that also the first case in Russia was documented in January?

The ambient temperatures could again provide the answer. Looking at the average temperatures in Moscow and Tokyo in the last three months, it is notably the difference between these two big metropolitan cities. I would expect that this gap will increase in the next weeks when the ambient temperatures will remain more favorable for the Japan.


Sweden and Greece

Table 6 – Summary table for Sweden and Greece. Top part of the table shows national data. Bottom part of the table shows local data for Stockholm and Athens (COVID-19 Cases refer to the Province or Region of the Capital city)

Sweden and Greece have very similar population and city´s sizes. In addition, both do not show high population density and these parameters explain why both countries are still in the “green zone” (Sweden is just above the limit) with respect to COVID-19 cases. However, the overall cases for Sweden are more than 3 times higher than those of Greece.

One explanation is that the first case in Sweden is dated 24th January 2020 while the first case in Greece is dated 26th February 2020. Thus, there is a gap of at least one month which can explain the lower number for the Greece. However, the population density in the Greek capital (3 times higher than Stockholm) should have speed up the spread. This did not happen.

Honestly, there might be several reasons. However, it is a fact that in the last two months (February and March), Athens is 10°C warmer than Stockholm, and this could have helped to slow down the contagion.


Italy and Republic of Korea (South Korea)

Table 7 – Summary table for Italy and Republic of Korea. Top part of the table shows national data. Bottom part of the table shows local data for Milan and Korea (COVID-19 Cases refer to the Province or Region of the Capital city)

This is a comparison that most of the Italian people will not like, since the name “Korea” brings to memory bad football moments (the defeat at the World Cup 2002). However, I thought it was worth to compare the two countries, basically because they have similar population and they are peninsulas. Moreover, at beginning of March, they were the most critical situations together with China.

Nowadays, the number of COVID-19 cases is extremely higher in Italy, with a steeper increase after the first week of March, while the increase in Korea flattened as shown few days ago in another article (

Following the previous comparisons, I investigated the ambient temperatures as well: there is almost a perfect match. In this case, the huge difference in COVID-19 cases cannot be explained by this factor. Honestly, I am not surprised. Positive results are the consequence of the combination of all the addressed factors, not just one.

Of course, I could have excluded this comparison from the analysis, but I thought that it was extremely important to present it. In fact, I am convinced (based on analyzed data shown in the other comparisons) that warmer temperatures can bring a positive outlook. However, they must be supported by the way we act daily. This cannot be ignored. In other words, this should be again a warning sign to keep respecting the social distancing rules and follow all precaution´s measures, making the most of the advantage that higher temperatures should bring in the next weeks.

To support what I believe and to complete the Italian analysis, I also made an internal comparison between Milan and Rome. Both areas are highly populated and with busy airport/transportation systems, even if the number of daily passages in the Milan area is definitively higher than Rome. Well, the COVID-19 cases are respectively 10819 and 2620. Looking at the average temperatures in the last two months, the gap between the two metropolitan areas is about 5°C, with Rome warmer than Milan. Assuming that the behavior of the inhabitants has been comparable, the slower rate of increase in Rome could be also attributed to more favorable temperature conditions.


United States of America (USA) and China

Table 8 – Summary table for USA and China. Top part of the table shows national data. Bottom part of the table shows local data for New York and Wuhan (COVID-19 Cases refer to the Province or Region of the Capital city)

The comparison is between the first nation where the crisis originated (China) and the nation (USA) where nowadays the most critical situation is, with more than 200000 cases and an increasing trendline that does not seem to slow down.

Both countries have a huge number of metropolitan cities, not comparable with any other countries. The general expectation is that when the virus initiates to spread then it is really hard to contain it. This is basically what the American situation shows in the last few days. This is not what the Chinese situation showed in the last two months. There are a lot of discussions about this special situation and the “real” documented cases. However, I do not want to speculate and I concentrate to other data that are independent from any possible manipulation or interpretation. One of them is the ambient temperature. Interestingly, simultaneously to the lockdown of the city, the temperatures in Wuhan increased in the last two months with peaks > 24°C and never below 4°C for the minimum temperatures. Thus, in addition to the isolation rule, quarantine and regular use of protective´s masks, the warmer ambient temperature might have helped to slow down the contagion.

On the hand, the city of New York is still “cold” with an average value comparable to the European countries and often still below 0°C. To support this consideration, looking inside the United States, it is interesting to compare the numbers in New York with those in Los Angeles: 114996 vs 13796. Basically 8:1 ratio. When we compare the ambient temperature in March 2020, the result is that New York is 8°C colder than Los Angeles (8°C vs 17°C), which shows peak temperature close to 30°C.

So, this comparison brings again to consider the relevance of the ambient temperature.


Portugal and Israel

Table 9 – Summary table for Portugal and Israel. Top part of the table shows national data. Bottom part of the table shows local data for Lisbon and Tel Aviv (COVID-19 Cases refer to the Province or Region of the Capital city)

When I looked at the COVID-19 data from these countries, I saw very similar results with slightly better situation for Israel. The population and the city´s sizes are not very different among the two countries: Israel has larger number of highly populated cities and higher population density. Moreover, the first registered cases were in both cases at beginning of March 2020. Therefore, my expectation was to see slightly worse results for Israel, and not for Portugal.

Moving to the local analysis and focusing on the ambient temperatures of Lisbon and Tel Aviv, there is a gap of about 3°C that could explain the slow down for the Israelian city. In particular, I decided to deepen the analysis and the last two weeks of March were definitively warmer with peak of about 30°C in Tel Aviv (highest temperature in Lisbon was 21°C in the same period but most of the time below 20°C).

I do not have any scientific proof that the higher temperature might reduce the contamination by COVID-19, but this is another interesting comparison to support this hypothesis.


France and the United Kingdom

Table 10 – Summary table for France and the United Kingdom (UK). Top part of the table shows national data. Bottom part of the table shows local data for Paris and London (COVID-19 Cases refer to the Province or Region of the Capital city)

Well, this is another interesting comparison. At moment, France has definitively higher number of COVID-19 cases compared to UK. The population and the city´s size are quite similar (at least the most populated) with a higher population density for the UK. Based on the latter, a larger case for UK was expected. This is not the case. The reason is most likely to be find on the fact that the situation for these two countries is primarily driven by the capital cities and their surroundings.

In fact, when we look at Paris and London data, we can observe that the huge difference is due to the higher population density of Paris versus the one in London. Therefore, an opposite trend to the national one. This aspect seems to justify the worse situation for the French capital. On the other hand, the absolute number of population and ambient temperature do not seem to play a significant role.

Of course, the relatively low temperatures for both countries could explain the daily high number of new cases. This is in line with other European countries.


Mexico and India

Table 11 – Summary table for Mexico and India. Top part of the table shows national data. Bottom part of the table shows local data for Mexico City and New Delhi (COVID-19 Cases refer to the Province or Region of the Capital city)

I decided to compare these two countries because they are really a special case. Both have very low COVID-19 cases compared to the population and compared to the large number of cities with more than 1 Million inhabitants. Both countries have documented the first case more than one month ago: end of January 2020 in India, end of February 2020 in Mexico. At the same time, these two countries have completely different cultural and living styles, as well as food and drinks.

How is it possible that the number of cases is so low? Of course, one important aspect is how many suspected persons have been checked. However, assuming that the approach is the same like in other countries, I found two possible reasons. The first is the average age of the population (not showed in the table), with both only about 6-7% of over 65 people. This is a very low value, not far from those for Asian countries like Philippines or Indonesia that are showing similar rates of COVID-19 cases. The second is that the ambient temperature is also playing an important role. Namely, in the last weeks, both capital cities (Mexico City and New Delhi) had average temperatures with peaks of 30°C or higher, which is similar to Buenos Aires and Singapore presented respectively in Table 1 and Table 4.


South Africa and Australia

Table 12 – Summary table for South Africa and Australia. Top part of the table shows national data. Bottom part of the table shows local data for Cape Town and Sidney (COVID-19 Cases refer to the Province or Region of the Capital city)

These are two countries from the Southern Hemisphere, which only recently showed an increase of cases. To be more precise: the first case in Australia in documented at end of January 2020. However, the number of cases remained low in the first weeks (29 total cases at March 1st). Instead, it started to increase faster in the last month. It might be a coincidence but March 2020 has been “colder” than the previous two months, with low temperatures around 15°C in the last weeks compared to 20°C in January and February 2020.

On the other hand, South Africa registered the first case at beginning of March 2020. Looking as well at the ambient temperature, the last week of the month showed a significant drop with low temperatures around 10°C. In parallel, in the same weeks, there was a significant increase of COVID-19 cases.

The situation requires to be monitored in the following weeks, when the ambient temperature is expected to further decrease. If the effect of temperature is relevant, further increase of cases is expected. It is also important to remark that these two countries, as well as the representative cities, have lower population density compared to the countries in the Northern Hemisphere. This could still help to slow down the contagion.


Brazil and Turkey

Table 13 – Summary table for Brazil and Turkey. Top part of the table shows national data. Bottom part of the table shows local data for Sao Paulo and Istanbul (COVID-19 Cases refer to the Province or Region of the Capital city)

This last comparison does not add more information to what already explained before. Namely, Brazil shows low number of COVID-19 cases compared to Turkey. This is theoretically in contradiction with the fact that Brazil is highly more populated than Turkey, especially for the presence of several cities above 1 Million and 100000 inhabitants. The local situation (Sao Paulo vs Istanbul) show the same theoretical contradiction.

The enigma can be somehow explained looking again at the ambient temperature of the last three months. Basically, Sao Paulo is 12-15°C warmer than Istanbul. This is another data to support the impact of external ambient temperature on the contagion´s spread.


General conclusions

All the collected data indicate that the ambient temperature plays a role in the contagion´ s spread: higher the temperature, lower the increase of COVID-19 cases.

The analysis of the Equatorial and Southern Hemisphere countries indicates that the greatest effect in slowing down the contagion´s spread could be obtained with ambient temperatures greater than 20°C.

The respect of social distancing and the discipline to follow all the precaution´s measures would allow to make the most of the advantage that the higher temperatures should bring in the next weeks.

The ambient temperature is one of the contributing factors, not the only one. Population, population density and city´s sizes cannot be ignored.

I hope that this analysis brings a moderate optimism for the upcoming weeks in those countries where the situation is particularly critical in these days. At the same, this analysis should be a warn to the Southern Hemisphere countries to adopt and increase as fast as possible all precaution´s measures (unless already done), currently applied in the Northern Hemisphere.

For any comments or clarifications, please do not hesitate to contact me.

Andrea De Filippo

This entry was posted in COVID-19 (Corona Virus) and tagged , , , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *