dcyphr | COVID-19 transmission in Mainland China is associated with temperature and humidity: A time-series analysis



This study examines how average temperature (AT) and average relative humidity (ART) relates to COVID-19 cases. This study considers 30 Chinese provinces. There was a strong relation between AT, ART and COVID-19. The correlation was negative. This means that higher AT and ART related to lower COVID-19 cases.

When ARH was 67-85.5%%, every 1 °C increase in the AT led to a decrease in the daily confirmed cases by 36-57%%. When AT was 5.04-8.2 °C, every 1%% increase in ARH led to a decrease in the daily confirmed cases by 11-22%%. These associations were not consistent throughout China.


In December 2019 there was an outbreak of COVID-19 in China. Typical symptoms included fever, dry cough and pneumonia. The disease could progress and cause death. As of 23 March 2020, there were 81603 confirmed cases. There were also 3276 deaths.

The disease can spread from cough and sneeze droplets. This means that the environment can affect the spread. This is because humidity and temperature can affect the survival of the virus. 

There is no prior investigation on the effect of the weather on COVID-19. The researchers used data from all Chinese provinces to investigate this. The researchers aim to observe the link between climate factors and COVID-19.

Materials and Methods

Study Area and Data

The researchers  looked at daily counts of confirmed cases in all Chinese provinces. They used data from the National Health Commission of People's Republic of China. The definition of a confirmed case has been published online. The researchers did not include Tibet as there was only 1 case.

The climate data was taken from "Weather Underground". The researchers investigated "health seeking behaviour" by looking at internet data from "Baidu".

Statistical analysis

COVID-19 had a 14-day incubation period. The researchers factored this into their calculations to make a moving average. This was compared to AT and ARH.

Sensitivity Analysis

The researchers performed a sensitivity analysis to verify the results. 


33453 cases occurred in Hubei. This accounts for 74.7%% of confirmed cases. The average AT was -16.96 to 19.3°C. The average ARH was 17.93 to 86.20%%. The number of cases in Hubei increased sharply after 20 Jan 2020. The temperature was usually around 6 °C. The humidity was usually above 70%%. For other provinces over 88%% of cases occurred between 23 Jan and 8 Feb 2020. ARH and AT varied by province.

There was a significant interaction between AT and ARH. The Baidu index (health seeking behaviour) was also significant. The provinces of Zhejiang, Shandong, Hebei, Jilin, and Gansu, showed significant interaction between AT and ARH.

AT showed an association with COVID-19 cases. Higher AT had lower number of cases. There was a similar relationship with ARH and COVID-19 cases. Higher ARH had lower number of cases. When ARH was 67-85.5%%, every 1 °C increase in the AT led to a decrease in the daily confirmed cases by 36-57%%. When AT was 5.04-8.2 °C, every 1%% increase in ARH led to a decrease in the daily confirmed cases by 11-22%%. 


This study shows that AT and ARH influence COVID-19 cases in Hubei and some other provinces. The association was not consistent. 

The study period was longer for Hubei. This means that the results are more stable. The data from Hubei was from before travel restriction were implemented. This means that Hubei most likely demonstrates the real effect of AT and ARH on COVID-19 cases. This showed that increased AT and ART led to fewer cases.

Other coronaviruses seem to follow similar patterns. One study said that the risk of SARS was 18 times higher on a cold day compared to a warm day. SARS and MERS are more stable in cold and dry conditions. Low temperature and humidity allows more particles to be in the air. This is ideal for virus spread.

This study found that low temperature and humidly can also lead to more COVID-19 cases. The exact mechanism of this is unclear. One possibility is that it causes human mucus membranes to dry and rupture. 

When estimating COVID-19 risk, climate factors should be considered. In spring there may be fewer cases of COVID-19 in China due to higher AT. Northern regions should be careful as AT and ARH remain low.

 The study does have limitations. The researchers could not identify Potential COVID-19 risk factors like socio-economic status. The imposed travel restrictions could affect the correlation found in other provinces. The climate applies only to the capital of each province. Provinces other than Hubei had a short study period with many imported cases.

In conclusion, climate factors affect COVID-19 spread. This is likely related to AT and ARH. The inconsistencies between provinces need to be studied further.