1.1 Recent studies about effect of weather on the incidence SARS-CoV–2
SARS-COV–2 is a relatively large virus with a size of about 1–2 nm from a group of enveloped virus family with a positive-sense single-stranded RNA. The virus is transmitted by coughing and sneezing through direct contact with the infected people (respiratory droplets) and infected surfaces (can survive for hours outside the host). SARS-CoV–2 general signs and symptoms are fever, cough and breath problems. In severe cases, infection evolves to pneumonia, serious respiratory problems and, on rare occasions, it can be deadly, mainly for older people (Harmooshi et al., 2020).
While influenza virus is linked to cold weather and peak of flu season is reached in the middle of winter, it is still a question how SARS-CoV–2 is linked to weather. Widespread testing and dependence between cooler countries in the north may cause, that number of confirmed SARS-CoV–2 cases between cooler northern and warmer-humid southern regions is different. However, more southern countries have realized large-scale testing and the number of positive SARS-CoV–2 cases (per capita) are lower than in northern climate (Bukhari and Jamel, 2020).
Sajadi et al. (2020) predict Potential Spread in connection with Seasonality for SARS-CoV–2, using climate data from cities with prominent spread of SARS-CoV–2. Authors predict where are the most likely higher risks of spreading of SARS-COV–2. To date, SARS-CoV–2, has created a significant community spread in big cities in the countries along a narrow west east regions crudely along the 30–50° N latitude with similar weather conditions (average temperatures of 5–11° C, absolute humidity (4–7 g/m3) and low specific (3–6 g/kg)). Significant predictors for spreading of a virus are mainly population proximity and interaction of people through travel.
Tang et al., 2009 have researched the survival of the airborne human coronavirus HCV/229E. This virus was used by authors as a representative or surrogate of respiratory coronary viruses and it was examined under various temperature and relative humidity conditions. Authors demonstrated that under such conditions, the virus has a half-life of 27 / 67 / only 3 hours at 30% / 50% / 80% humidity. Here is an evidence, that near higher relative humidity, half-life of virus has considerably shorter time of survival. When the temperature drops as 6 °C, HCV/229E half-life is redoubled to 6 hours at 80% humidity. Similar results have been found for seasonal flu (Lowen and Steel, 2014). It is assumed that this severe effect on prolonging the virus’s half-life is linked to the spreading of the SARS-CoV–2 during low-temperature and high-humidity pattern of weather (Harmooshi et al., 2020).
Wang et al. (2020) researched the influence of air temperature and humidity on the transmission of SARS-CoV–2. Meteorological data for 100 Chinese cities with more than 40 cases were used. According to the study, the arrival of humid summer season could significantly diminish the transmission of the SARS-CoV–2.
Ma at al. (2020) explored the association between SARS-CoV–2 deaths and weather parameters (temperature and humidity) in 2299 SARS-CoV–2 related deaths. Both meteorological elements are considered as important factors for increasing/decreasing SARS-COV–2 mortality.
Chan et al. (2020) investigated on the effect of temperature and humidity on SARS-CoV–2. Authors used as target group of viruses HKU39849 strain. Facilitations of SARS-CoV–2 transmission were linked with lower temperature and dry conditions.
Xiao et al. (2020) found that big air pollution also has an effect on the impacts of SARS-CoV–2. An increase of only 1 μg/m3 in PM2.5 brings a 15% increase in the SARS-CoV–2 death rate at the level of significance 0.05. Results are statistically significant and robust to secondary and sensitivity analyses.
Using data on local transmissions until the 23rd of March 2020, Araujo and Naimi (2020) in a preprint article developed an ensemble of 200 ecological niche models to project monthly variation in climate suitability for spread of SARS-CoV–2 throughout a typical climatological year. Although cases of SARS-CoV–2 are reported all over the world, most outbreaks demonstrate a pattern of clustering in cooler and drier regions. The predecessor SARS-CoV–1 was linked to similar—dry and cold climate conditions. Authors expect asynchronous seasonal global outbreaks unless current trends of spreading SARS CoV–2 will be continuing. Models have shown that temperate cold and humid climates are more favorable to spread of the virus, whereas tropical hot and arid climates are less favorable. Nevertheless, model uncertainties are still high in the regions of sub-Saharan Africa, Latin America and South East Asia. Unsuitable climates can cause quick destabilization and reducing capacity of the virus.
Correspondence of Brassey et al. (2020) suggests that cold and dry conditions may influence the transmission of SARS-CoV–2. Authors gave a summary of evidences from unpublished papers till 22. March 2020:
- A cross-sectional study found that every 1℃ of increase in the minimum temperature led to a decrease in the cumulative number of cases by 0.86.
- A modeling study suggested a transitory reduction of incidence in spring and summer period and subsequently increase during the winter period 2020–2021.
- Another modeling study found that the current spread preferences are cool and dry conditions. 2002–3 SARS-CoV–1 outbreak was linked to similar weather conditions.
Conclusions and predictions of these studies have high uncertainty (their results are potentially biased because of uncontrolled confounding).
- Next two studies announce that daily mortality of SARS-CoV–2 is positively correlated with diurnal temperature range (DTR) but negatively correlated with relative humidity. Temperature and relative humidity are likely to contribute a maximum of 18% of the variation in transmission. Every 1° C increase in temperature / 1% increase in relative humidity lowered the R by 0.0383 / 0.0224.
- A very recent modeling study investigated the relationship between temperature and predicted the number of cases. Lower temperatures were worse for incidence of SARS-CoV–2 than were higher ones.
Lin et al. (2006) investigated to identify factors involved in the emergence, prevention and elimination of SARS-CoV–1 in Hong Kong from 11 March to 22 May 2003. A structured multiphase regression analysis has shown the potential effects of weather, time and interaction effect of hospital infection on severe acute respiratory syndrome. In colder days during the epidemic, the risk of increased daily incidence of SARS-CoV–1 was 18.18-fold (5.6–58.8 at the level of statistical significance 0.05) higher than in warmer days. Naturally decrease during the epidemic might by an average of 2.8 the total daily new cases every 10 days. The authors considered SARS-CoV–1 transmission to be dependent on temperature changes during the climate year and on the multiplicative effect of hospital infection. SARS-CoV–1 retreated naturally during the further development of the epidemic.
Tan et al. (2005) have found a significant correlation between the SARS-CoV–1 cases and the air temperature seven days before the incidence and the seven-day time lag corresponded well with the well-known incubation period. The optimum air temperature linked with the SARS-CoV–1 cases (and its encourage virus growth) was from 16 to 28° C. A sharp decrease (and too rise) of air temperatures related to the cold spell led to an increase of the SARS-CoV–1 cases - human immune system is influenced by cold weather. There are some evidences that a higher possibility for SARS-CoV–1 to reoccur is during the spring more than the autumn / winter.
Gardner et al. (2019) studied a background of Middle East respiratory syndrome coronavirus (MERS-CoV) from January 2015 to December 2017. To find linkages between primary MERS-CoV cases and previous weather conditions within the incubation period (2 weeks) in Saudi Arabia a case-crossover design using invariable conditional logistic regression was used. The full case dataset (1191 cases) was reduced to representative group most likely to represent spillover transmission from camels (446 cases). In research, meteorological data from localities closest to the largest city for each Saudi region were used (daily maximum, minimum and mean temperature (in ° C), relative humidity (in %), wind speed (in m/s), and visibility (in m). Lowest temperature (Odds Ratio = 1.27; 1.04–1.56 at the level of statistical significance 0.05) and humidity (Odds Ratio = 1.35; 1.10–1.65 at the level of statistical significance 0.05) were linked with incidence of MERS-CoV 8–10 days later, high visibility 7 days later (Odds Ratio = 1.26; 1.01–1.57 at the level of statistical significance 0.05) and wind speed 5–6 days later.
Given the previous associations between viral transmission and humidity and temperature across which the majority of the SARS-CoV–2 cases have been observed until the present date, that lower humidity and lower temperatures are unfavorable for incidence of SARS-CoV–2. If SARS-CoV–2 is relatively sensitive to these types of weather, then it could be applied to optimize the SARS-CoV–2 mitigation strategies (Bukhari and Jamel, 2020). The provision of protective gear is also a very important factor for the prevention of SARS-CoV–2 infection. Other circulating viruses, such as flu copy a seasonal effect, and for that reason co-infection rates will have decline trend, which may have effect on death rate (Brassey et al., 2020).
1.2 Main circulation patterns in the mid-latitudes (North Atlantic Oscillation/NAO and Arctic Oscillation/AO phases)
Seasonal climatic anomalies manifest themselves over large geographical regions. Some areas appear cooler and drier, whereas others, hundreds to thousands of kilometers away, warmer and wetter. These parallel structures with opposite phases and manifestations in climatology are referred to as teleconnections. Most often, they are determined using one point correlation maps, comparing the correlations of the height of the grid field points of the pressure level of 500 hPa and the reference point. Low-frequency oscillations are also reflected in the ground pressure field. After depicting air pressure anomalies at sea level, the ground air pressure dipole is visible (Hurrell et al., 2003). A total of 13 modes have been identified that affect the northern hemisphere climate. The NAO (AO) mode is the only one of all modes having expression in Europe throughout the year, the other 3 main modes - EA, EU2 and EU1, are only visible in selected seasons. Among teleconnections we can include Southern Oscillation (ENSO) too, but it has no significant manifestations in Europe, it is considered with AO to be the most extensive oscillation affecting the climate on Earth.
The magnitude of the pressure gradient between the Arctic and moderate latitudes is closely related to the evolution of the polar vortex and the nature of the jet stream. In meteorology, this magnitude is linked with two major oscillation modes occurring above the northern hemisphere: Arctic Oscillation (AO) and North Atlantic Oscillation (NAO). AO is clearly identifiable in pressure fields (especially in winter) throughout the northern hemisphere, while the NAO is its equivalent to the Euro-Atlantic area. 90% of the oscillations are interconnected. If the AO and NAO indices both show negative values (negative phase), this results in a weaker and less pronounced polar vortex and a weaker and meandering jet stream. Stronger polar vortex and more zonal running and strong jet the stream, on the other hand, is associated with the positive phase of AO and NAO. Each phase has its typical manifestations in pressure, temperature and precipitation fields in individual regions, including Europe (Hurrell, 1995, Hurrell and Desser, 2009; Hurrell a van Loon, 1997; Kapala et al., 1998; Wanner et al., 2001).
The NAO describes in a meridional direction the transfer of atmospheric masses between the subtropical region in the Azores Pressure Area and the sub-polar in the Icelandic Pressure Area. It affects both seasonal and interdecadal variability in global circulation (Hurrell, 1995) and is the only oscillating mode to show statistically significant links with meteorological characteristics in Europe throughout the year. According to Hurrell et al. (2003) the area of influence of the NAO extends from the east of the USA to Siberia in the zonal and from the Arctic to the subtropical Atlantic in the meridional direction.
There are two types of situations: type A - equivalent of negative phase: NAO-/AO- (weakening of pressure formations in the North Atlantic) and type B - equivalent of positive phase: NAO+/AO+ (more developed pressure units). The distribution of pressure units over the Atlantic varies considerably in winter and summer - in the summer, the Azores High (the first NAO Action Center) dominates, which also extends northwards due to the shift of the intertropical convergence zone to the north. On the contrary, it weakens and shifts to the south during winter, with Icelandic Low dominating the Euro-Atlantic area (second NAO action center). Differences in air pressure deviations between the positive and negative NAO phases in action center areas are over 15 hPa. Significant changes in atmospheric circulation remodeling occur when the NAO index exceeds +1 or –1 (Hurrell, 1995).
During the positive phase of the NAO, anomalously higher air pressure occurs west to southwest of the coast of Europe in combination with anomalously low air pressure in the Arctic Atlantic. The flow generated by the rotation of pressure formations is intensified by a larger pressure gradient, causing a more pronounced western flow at moderate latitudes over (North) Atlantic and (Northern) Europe, a southern flow over the east coast of the US and north over western Greenland and Canada. The western flow brings warmer and damper sea air to Europe in winter and causes milder winter, while the northern flow from the Great Lakes to the north and northeast causes more waves of cold and blizzards. In the negative phase, the situation is reversed: the pressure formations above Iceland and the Azores are anomalously weakened (shallower Iceland low and less robust Azores high). A weak meridional pressure gradient causes a weak western flow, so much humid and warm ocean air does not get above Europe in winter, a suitable situation for meridional air mass transfer occurs and potential invasions of Arctic and Siberian air. With the extremely low oscillation index, seasonal Greenland High increases, bringing above-average air temperatures in Canada, the United States and Greenland, but significant cold air inflows over Europe. The following patterns of correlation of the NAO index with the most important meteorological elements - precipitation, air temperature and pressure field distribution (Blade et al., 2012) were identified:
1.) Summer: During the positive phase of the NAO, there is a robust pressure high above the British Isles, extending from the Azores to Scandinavia, with a center in Scotland and Northern Ireland. There is a deep low-pressure area above Greenland, and a low-pressure field extends over the Mediterranean. On the contrary, during the negative phase of the NAO, anomalous high-pressure spreads over the area of Iceland and Greenland, and areas of mid- and subtropical latitudes from North America over the Atlantic to mid-latitudes of Europe and the Mediterranean report negative surface pressure anomalies. The largest negative pressure anomalies are achieved in northern Germany and Denmark (Boé et al., 2008). Positive correlations with precipitation are achieved during the summer and positive NAO phases in the British Isles, Benelux, north of France, Germany and Poland to the Eastern Carpathians, the south of Scandinavia and the Baltic Sea. The robust high pressure above Northwest Europe during the NAO + (Action Center) causes a flow of cooler and drier air over the continent, which has a stabilizing effect on temperatures and precipitation in Central Europe. Negative correlations with precipitation have been confirmed across the Mediterranean, from Spain to Turkey, with the highest values in the Adriatic and the Balkan. The penetration of cooler air from the northwest generates numerous precipitations and low temperatures - mainly above the Balkans, Italy, Turkey, to Israel and Egypt. On the contrary, the robust pressure above the British Isles brings, in addition to drought, above-average temperatures on the British Isles, and from west and north-west coasts of Europe to the Baltic. During the negative NAO phase, temperature and precipitation responses are reversed.
2.) Winter: The different nature of winters is largely related to the tracking of Atlantic cyclones in winter. During the winter when the polar vortex is locked above the sub-polar region and the zonal flow reaches high levels, the low-pressure action center is located more northerly on the Greenland - Iceland axis. The deep Icelandic low travel across Europe by northern tracks - over the British Isles and Scandinavia, at most the coast of the North Sea and the Baltic Sea. Europe (except the British Isles and Scandinavia) is on the front of the quasi-stationary deep low pressure zone in the warm, predominantly southwestern, during the winter NAO + phases, while stormtrack is shifting from the Island-British Isles axis to the north of Europe. The activity of Mediterranean cyclones is minimal - in the Mediterranean there is dry and relatively cold weather (thanks to clear anticyclonic nights). The northernmost and, in particular, the more robust regions of the Azores high reach western, southern but also central Europe more often than usual during these phases. During the negative phase the situation is reversed - the stormtracks are due to the extensive meanders of the polar and arctic front and weakened zonal air transport curved over the Mediterranean. Due to the disintegration of the polar vortex, the jet-stream is forced to the south, generating low pressure over lower latitudes (southern, moderate and subtropical regions). The low pressure over the Mediterranean and the high pressure over Scandinavia and the northern half of Europe, together with the more pronounced Siberian anticyclone, then create suitable conditions for invasions of Arctic or dry and cold Siberian air over Europe (the Mediterranean pressure lows “pulling” cold air towards them). Pressure field such this, typical for NAO-phases, is associated with cyclonic weather with higher temperatures and higher rainfall totals over the Mediterranean, and drier and cold conditions over the northern half of Europe. In winter, correlations of the NAO index with temperatures and precipitation have a stronger zonal pattern than in summer. A clear signature of the positive NAO phase on air temperatures, precipitation, as well as the total pressure field and flow is present from North America to eastern Siberia and is associated with overall warming in Siberia and cooling in eastern Canada and from subtropical to tropical latitudes.