Lightning Behaviour during the COVID-19 Pandemic

Background COVID-19 has drastically dampened human activities since early 2020. Studies have shown that this has resulted in changes in air temperature and humidity. Since lightning activities are dependent on air temperature and humidity, this study is conducted to evaluate the correlation between the intensity of lightning activities with the atmospheric changes, and investigates the changes, in lightning activities due to atmospheric changes during the COVID-19 pandemic. Methods The hypothesis was tested through a t-test and Pearson’s correlation study. The variation trend of lightning strikes count (LSC) in Europe and Oceania during the five months COVID-19 lockdown period (March – July) compared to the same period in the previous five years from 2015 to 2019 is investigated. Results Statistical analysis shows the LSC in Europe and Oceania during the lockdown period dropped significantly by more than 50% and 44% respectively compared to the same period in previous five years. Furthermore, LSC was found to be positively correlated with air temperature and relative humidity in Europe. However, in Oceania, LSC seems to be only positively correlated with air temperature but negatively correlated with relative humidity. Conclusions This study seems to suggest that lightning activities have significantly changed during this pandemic due to reduction in human activities.


Introduction
Many countries have enforced lockdown since the beginning of the COVID-19 pandemic. [1][2][3] Energy-intensive human activities such as travelling and the hospitality sector were drastically reduced resulting in reduced emissions of greenhouse gases. 4 The global CO 2 emission is estimated to drop by 8.8% (À1551 Mt) in the first half-year of 2020 compared to the same period in 2019. Moreover, almost 18% of CO 2 emissions in recent years were produced from ground transportation. 5 With the exception of the preliminary findings by Jones et al., 6 the general expectation by researchers is that the trend of temperature is expected to be reduced due to the reduction in CO 2 . A significant positive correlation between the atmospheric temperature and CO 2 emission is reported in. 7 Furthermore, COVID-19 lockdown has caused micro-climate changes such as localized variations in air temperature and relative humidity. 8 The pandemic is also having an effect on NO X , causing a decline that could possibly lead to short-term cooling. 9 Air humidity will also be affected as global warming are dependent on both temperature and humidity. 10 This reduction in human activities could also result in drop in aerosol level globally. This reduction in human activities could also result in drop in aerosol level globally. The result of lockdown has disrupted human and industrial activities around the world. The lockdown generally leads to a notable change in carbon dioxide (CO 2 ), temperature, and humidity. However, the reduction in human activities may reduce greenhouse gases and may result in a drop in global temperature. The study by Singh et al. found that implementation of lockdown reduced the percentage of temperature and may mitigate the pace of climate change in the future. 8 Lightning, a natural atmospheric discharge, is affected by various environmental factors. Lightning brings about hazards to human life and appropriate risk assessment has to be conducted for any habitable structure. 11,12 Atmospheric variables such as climate change, humidity, aerosol level, and wind motion can affect the cloud charge distribution, electric field and threshold electromagnetic fields that give rise to air breakdown. It is predicted that lightning may strike more frequently as a result of the ongoing climate change. 13 The lightning intensity may also increase due to the high greenhouse gases in the atmosphere. However, study by Finney et al. stated that many previous studies found a positive correlation between lightning and temperature, and one previous study that found lightning decreases with an increase in temperature. 14 This may explain that such relationships become highly uncertain on longer timescales. When warm, wet air rises into the cold air, thunderstorms form. As the warm air cools, moisture in the form of water vapour condenses into water droplets, a process known as condensation. Cooled air descends through the atmosphere, warms up, and rises again. A convection cell is a circuit of rising and descending air. A cloud will form if this happens in a small amount. A thunderstorm can arise if this happens with a lot of air and moisture. The presence of high air temperature and high relative humidity can quickly rise and cause powerful updrafts. These updrafts carried water droplets and quickly froze and collided with ice crystals and graupel, causing the charge transfer process.
Lightning could also be triggered by aerosols released by industrial processes and transportation activities. 15,16 Aerosol could affect lightning activity through modification of cloud micro-physics. Aerosol particles serve as cloud condensation nuclei and ice nuclei, and the amount of this particles could affect the formation of cloud droplets and ice particles. More aerosol will suppress the coalescence and making the average size of cloud droplet to be reduced as well as inhibiting precipitation. Therefore, the process enables the water droplets to rise further to upper layers of the clouds and may enhance the lightning processes. During the lockdown period, many industrial sectors stopped operating. Thus, human activities have considerably reduced during the COVID-19 pandemic which may affect the rate of lightning. Lightning ground flash density tends to increase with drier and warmer surface air. 17 Furthermore, the frequency of thunderstorms shows a major peak during summer time. 19 Previous studies have also found a strong relationship between relative humidity and lightning occurrence. 19

REVISED Amendments from Version 2
We have added the following sentence to clarify our assumption of Detection Efficiency: "It is assumed that the Detection Efficiency of the sensors used by LightningMaps.org remained constant during the period considered in this study." Any further responses from the reviewers can be found at the end of the article Hence, it is of interest to investigate the correlation between the environmental changes that happened during the period of COVID-19 related restriction of human activities and the lightning occurrence density. This study is an attempt to analyse this situation. This study investigates the trend of five months of lightning occurring from March to July in 2020 compared with the same period (March-July) in 2015-2019 in Europe and Oceania. The outcomes of this work could yield interesting insights into the correlation between human activities and lightning frequency.

Overview
Lightning stroke counts (LSC) and two atmospheric factors namely air temperature and relative humidity are considered as the variables in this study. The relationship between LSC with respect to air temperature and relative humidity will be statistically analysed via the dependent t-test and Pearson correlational studies.  Tables 3 and 4 show the average value of air temperature and relative humidity in Europe and Oceania in year 2020.

Statistical approach
A dependent t-test is was conducted using Microsoft Excel 2016 (Microsoft Excel, RRID:SCR_016137) to determine whether there is a statistically significant difference between the LSC during the lockdown period in the year 2020 and the LSC in the same period (March-July) in year 2015 until 2019. The LSC is measured from a single population (Europe or Oceania) and two different timelines (before and during). Period A represents the lightning activities before lockdown   The confidence level of 95% at a significant level, α ¼ 0:05 is used. This approach tests the hypothesis and calculates the probability of determining whether there is evidence to reject the null hypothesis. When the P value < 0.05, the null hypothesis is rejected, and vice versa.
Next, the Pearson correlation coefficient is used to evaluate the correlation between the frequency of lightning activities with the atmospheric changes. The Pearson's correlation coefficient, r, is computed to measure the strength of the relationship between total lightning strikes, air temperature, and relative humidity in Period B.
Furthermore, the correlation between the variables was analysed using regression and correlation analyses. The significant level, P value can be obtained from the regression data analysis. The null hypothesis, H 0 and the alternative hypothesis, H a is defined as below: Null hypothesis, H 0 : P = 0, There is no significant relationship between lightning strikes with air temperature or relative humidity.
Alternative hypothesis, H A : P 6 ¼ 0, There is a significant relationship between lightning strikes with air temperature or relative humidity By using the P-value method (α ¼ 0:05), the decision on rejection or acceptance of the null hypothesis can be made. There is sufficient evidence to conclude that there is a significant correlation between lightning strikes and air temperature or relative humidity as the correlation coefficient is significantly different from zero. Exact P values and the mean value of lightning strikes from May to July are provided in Table 5.

Results and discussion
Europe Figure 1 shows the LSC has dropped significantly in the year 2020 when the lockdown started. The dependent t-test shows a statistically significant (P-value <0.05) difference between 2020 and each previous year as shown in Table 5.
Notably, LSC in Europe during the five-month lockdown period were reduced by more than 50% compared to the same period in the year 2019, 2018, and 2017. Figure 2 illustrates the variation of LSC against air temperature levels in Europe. Figure 3. illustrates the relationship between LSC and relative humidity in Europe. Table 6 shows that the correlation of lightning strikes with air temperature and relative humidity in Europe are statistically significant. The Pearson correlation between lightning strikes and air      There was a 44% drop in LSC from 2019 to 2020 as shown in Figure 4. Table 7 shows there is statistically significant difference between the year 2020 with all previous years except 2017. Figure 5 and Table 8 indicates a moderate positive correlation between LSC and air temperature in Oceania during the lockdown period. Unlike Europe, Figure 6 and Table 8 shows that the relationship between LSC and relative humidity in Oceania is negatively correlated. The positive correlation of LSC and air temperature is consistent with previous studies. 28,29 The negative correlation of LSC and relative humidity in Oceania obtained in this study contradicted the study of Shi et al. 20

Conclusions
In conclusion, there was a drastic drop in LSC in Europe and Oceania during the first lockdown period in 2020. A dependent t-test confirmed that a statistically significant difference in LSC between Period A and Period B. There is a positive relationship between LSC and air temperature in Europe (r = 0.92) and Oceania (r = 0.55). Furthermore, there is a positive relationship between LSC and relative humidity in Europe (r = 0.52) but a negative relationship between LSC and relative humidity in Oceania (r = À0.54).
The difference in correlation findings between lightning and relative humidity in Europe and Oceania remains unexplained. Higher relative humidity will lead to stronger updraft and increased lightning occurrence. However, too much vapor may weaken the updraft by blocking the vapor to rise up to complete the phase transformation.
The differences in correlation between lightning, air temperature, and relative humidity in Europe and Oceania may also be due to other possible factors such as aerosol level, wind motions, and particulate matter. Future work should be replicated in other geographical regions such as America and Asia.

Comment:
This is an interesting study on the lightning activities during the covid-19 pandemic periods. This article also relates lightning to air temperature and humidity. However, there are some parts of the article that can be further justified in order to improve the paper quality. Some suggestions have been made here: What is the definition of the mean value shown in table 5?

Response:
Thank you for the comment. It actually means the mean value of lightning strikes from March-July. We have added the following to clarify: Exact P values and the mean value of lightning strikes from May to July are provided in Table 5 Comment: Perhaps the authors can further explain the relationship between temperature and LSC. Why does low temperature mean less LSC etc?

Response:
Thank you for the comment. We have added the following explanation in paragraph 3: When warm, wet air rises into cold air, thunderstorms form. As the warm air cools, moisture in the form of water vapour condenses into water droplets, a process known as condensation.
Cooled air descends through the atmosphere, warms up, and rises again. A convection cell is a circuit of rising and descending air. A cloud will form if this happens in a small amount. A thunderstorm can arise if this happens with a lot of air and moisture. The presence of high air temperature can quickly rise and cause powerful updrafts. These updrafts carried water droplets and quickly freeze and begin collision with ice crystals and graupel which causing charge transfer process. Therefore, low temperature may slow down this process.

Comment:
Perhaps the authors can further explain the relationship between humidity and LSC. Why does low temperature mean less LSC etc?

Response:
Thank you for the comment. Studies found that higher relative humidity may enhance the upward updraft and easing the particle collision in the cloud. Lower relative humidity may lead to weaker updraft and decreased the chance of lightning occurrence. This trend has been demonstrated in Europe but not in Oceania. We have added the explanation in the Results and Discussion section under Europe.

Comment:
A suggestion for the authors to discuss the relationship between months and lightning events/LSC, whether the lockdown or economic slow down affects the lightning occurrence. The authors may provide reasons to justify it as well.

Introduction
Second paragraph: Please mention the reduction in aerosols during the lockdown.
The reference provided by the authors for the relationship between CO2 emissions and temperature reports a significant positive correlation between the atmospheric temperature and CO2 emission at a scale of years. Please provide some reference for a possible correlation between CO2 and temperature at a time period similar to the lockdown (weeks-months). Jones et al. (2021) 1 did not find any relationship between the CO2 reduction and changes in temperature.  5 .

Methods
The authors have to estimate the temporal evolution of the lightning Detection Efficiency of the sensors used by LightningMaps.org, as you will study the temporal evolution of lightning in the data set. Lightning can increase or decrease year by year due to changes in the Detection Efficiency. The Detection Efficiency of the data set can be investigated by comparison with other lightning data set, such as LIS.

General comments and conclusions
This study shows that there could be a possible relationship between temperature, relative humidity and lightning in the studied regions (Europe and Oceania). However, the influence of the COVID-19 pandemic is not clear. There were significant changes in both temperature and relative humidity during the pandemic that could influence lightning. However, such changes could not be connected to the pandemic. See for example Jones et al. (2021) 1 , who used 12 models to produced over 300 simulations. They did not find any associated impact of the reduction of CO2 on temperature or rainfall. Changes observed in temperature and relative humidity can be due to other factors instead of the pandemic.
In addition, the authors find an opposite correlation between relative humidity and lightning in The release of carbon dioxide (CO2) from the combustion of coal, oil, and natural gas is the primary driver of rising global temperatures. These activities, as can be seen, tend to decrease during the lockdown period. As a result, the rate of temperature rise may be delayed during the lockdown period. This, however, is insufficient to cause long-term effects on CO2 levels and climate change on a worldwide scale.

Comment:
Third paragraph: Some authors have reported a possible decrease in lightning activity due to climate change (Finney et al., 2018).

Response:
Thank you for the comment. The recommended reference has been cited. Here is our response: (Finney et al., 2018) stated that many previous studies found a positive correlation between lightning and temperature, and one previous study that found lightning decreases with an increase in temperature. This may explain that such relationships become highly uncertain on longer timescales.

Comment:
Fourth paragraph: Please explain how aerosols are involved in lightning activity.

Response:
Thank you for the comment. The following has been added into fourth paragraph: Aerosol could affect lightning activity through modification of cloud micro-physics. Aerosol particles serve as cloud condensation nuclei and ice nuclei, and the amount of this particles could affect the formation of cloud droplets and ice particles. More aerosol will suppress the coalescence and making the average size of cloud droplet to be reduced as well as inhibiting precipitation. Therefore, the process enables the water droplets to rise further to upper layers of the clouds and may enhance the lightning processes.

Comment:
"Lightning ground flash density tends to increase with drier and warmer surface air": Please provide some reference. In general, lightning does not tend to increase with drier surface.

Response:
Thank you for the comment. We have provided reference based on the study from Diaz-Ortiz reference number 17 into third paragraph: Yes, in general lightning does not tend to increase with warmer and drier surface, but the presence of high air temperature can quickly rise and cause powerful updrafts. These updrafts carried water droplets and quickly freeze and begin collision with ice crystals and graupel which causing charge transfer process. Therefore, low temperature may slow down this process. So high temperature can be a driving factor to lightning formation. Comment: