COVID-19 and The Impact of Climatic Parameters: A Case Study of Bangladesh

This study examines the relationship between in Bangladesh. Pearson correlation coefficient, Spearman correlation coefficient, and Kendall's correlation coefficient have all been put to use to assess the intensity and direction of the relationship between climatic factors and COVID-19. The lagged effects of climatic parameters on COVID-19 daily-confirmed cases from Bangladesh are being looked into using the Auto Regressive Distributed Lag (ARDL) model. As a result, two non-climatic variables, such as population density and the human development index, are taken into account as control variables. As climatic variables, average temperature (°C), average humidity (percent), average PM 2.5, and average wind speed (km/h) were well chosen. The time series data used in this analysis was from May 1, 2020 to April 14, 2021. The findings of correlation analysis indicate that there is an important, significant, and positive relationship between COVID-19 widespread and temperature (°C), humidity (percent), and wind speed (km/h), whereas there is a negative, weak, and significant relationship between PM 2.5 and COVID-19 widespread. In addition, the ARDL findings suggested that temperature (°C), PM 2.5, and wind speed (km/h) have major lagged effects on COVID-19 in Bangladesh, while humidity (percent) has negligible lagged effects. For policymakers and investors alike, the consequences of this study are important in Bangladesh.


Introduction
COVID-19 disease was first detected in Wuhan, China, in December of last year. COVID-19 is a new form of virus that spreads quickly from person to person and causes a large-scale outbreak.
The novel Corona virus has infected people in nearly every country on the planet. The disease's structure is very complex, and it spreads quickly. As a result, understanding disease patterns has become critical. Sadly, as of May 8, 2021, 157 million COVID-19 cases had been confirmed worldwide, with 3.27 million deaths (WHO). Since no treatment method for this virus has been identified, effective design of the health foundation and structure, where the figure of disease lay out should be commended, is required. It is important to track and forecast status in order to better manages the spread of this disease.
Bangladesh is a small nation with a relatively stable population. Bangladesh's weather is unprotected (Islam 2019). Bangladesh is the third country in South Asia to be affected, following India and Pakistan. On March 8, 2020, Bangladesh reported the first COVID-19 cases (IEDCR).
Dhaka, Bangladesh's capital is the most severely affected city. On June 13, 2020, Bangladesh outnumbers China in terms of COVID-19 confirmed cases. Bangladesh's government announced a 10-day curfew on March 23, 2020, which was later extended to May 30, 2021, in  Several experiments have been performed to see whether the corona virus can be transmitted by environmental factors. In humid focused areas, Pramanik et al. (2020) discovered a close association between temperature and COVID-19 events. Some studies have found that certain meteorological variables, such as temperature, PM 2.5, and median age, are important factors in the spread of the corona virus Gupta et al., 2020;Islam et al., 2020;Prata et al., 2020;Espejo et al., 2020;Jamil et al., 2020;Qi et al., 2020;Wang et al., 2020). Dogan et al.
(2020) used correlation analysis and the ARDL model to investigate the impact of meteorological factors on COVID-19. COVID-19 appears to have a negative relationship with temperature, while COVID-19 appears to have a positive relationship with humidity and air quality. COVID-19 is also affected by meteorological conditions that are delayed. Cai et al. (2020) investigated the relationship between temperature and COVID-19 mortality and discovered that COVID-19 mortality and temperature are positively correlated, while COVID-19 mortality and humidity are negatively correlated.
To examine the relationship between climatic parameters and COVID-19, Islam et al. (2021) used a single study. There is no association between 7days lagged climatic factors and COVID-19, according to the author's findings. Although there is a positive relationship between COVID-19 and 14days lagged temperature, there is a negative relationship between COVID-19 and 14days lagged humidity. Duhon et al. (2021) used a multiple regression model that included climate, social, demographic, and non-pharmaceutical variables, and found that demographic and social factors are to blame for COVID-19 spread.

Page 6
Abbreviations: GAM, Generalized Additive Model; SARS, Severe Acute Respiratory Syndrome Chen et al. (2020) discovered a close connection between COVID-19 transmission and meteorological parameters. COVID-19 has a negative relationship with wind speed and ventilation coefficient, but a positive relationship with humidity, dew point, and temperature, according to Pani S K et al. (2020). Yuan et al. (2021) used the GAM model to look at the effects of climatic factors on COVID-19, and discovered that climatic factors are nonlinearly related to COVID-19. Many diseases are spreadable can spread via epidemiological channels (Shi et al., 2020;Wang et al., 2020). COVID-19 distribution could be influenced by changes in temperature, air quality, and wind speed (Sobral et al., 2020;Shakil et al., 2020;Bashir et al., 2020a;Suhaimi et al., 2020;Kumar S. 2020;Lorenzo et al., 2021;Bashir et al., 2020b).
The COVID-19 growth rate has a clear and optimistic relationship with emissions. As a result, it's critical to look at how climatic and weather conditions are linked to COVID-19 transmission for the sake of human survival and well-being. Coskun et al. (2021) discovered a near link between COVID-19 widespread and wind speed. Wind or air circulation can spread COVID-19.
Wind and population density were significant determinants in virus propagation, accounting for 94% of the variation. Population density completely mediated the wind effect on virus propagation according to author findings.
According to some studies, the SARS virus was also transmitted by weather conditions (Tan et al., 2005;Yip et al., 2007;Yuan et al., 2006). Cai et al. (2007) used logistic regression analysis to investigate the relationship between meteorological factors and human health (average temperature, sunshine, humidity, and air pressure). Abbreviations: SARS-CoV-2, severe acute respiratory syndrome corona virus 2 The transmission of the SARS virus is discovered to be inextricably connected to meteorological influences, according to the authors. SARS spread quickly, according to Yuan et al. (2006), due to three main meteorological factors: temperature, wind velocity, and relative humidity. At this point, it was discovered that certain climatic and weather factors could affect influenza transmission (Chen et al., 2017;Bai et al., 2019;Kalhori et al., 2019 SARS-CoV-2 has recently been reported to be spreadable through airborne transmission (Prather et al., 2020;Hadei et al., 2020;Allen et al., 2020;Klompas et al., 2020;Greenhalgh et al., 2021;Ahlawat et al., 2020;Ueki et al., 2020). Morawska and Cao (2020) discovered a close connection between SARS-CoV-2 transmission and airborne transmission. COVID-19 will spread through the air, according to Yao et al. (2020), and it is negatively correlated with higher ozone levels.
They also suggest that environmental factors and social isolation could be contributing to the pandemic's spread. SARS-CoV-2 could spread through airborne transmission, according to Correia et al. (2020).

Page 8
Weather conditions, temperature, and relative humidity have been linked to COVID-19 in some studies (Briz-Redón et al., 2020;Adhikari et al., 2020;Chien et al., 2020).  transmission is caused by comparatively dry and cold weather, according to Fu et al., (2021). Tosepu et al., (2020) assume that climatic parameters and COVID-19 spread are inextricably linked. According to Pan et al. (2021), there is no connection between weather factors and COVID-19 transmissions, and that warm weather can reduce COVID-19 transmission. Wu et al. (2020) used the GAM model to investigate the relationship between COVID-19 and meteorological influences, finding that relative humidity, wind speed, and temperature have nonlinear effects on COVID-19 regular new cases when temperature, relative humidity and wind speed were below 20°C, 70% and 7 m/s respectively. According to Leal and Hernández (2020), temperature has a negative relationship with COVID-19, and COVID-19 cases with carbon monoxide have a clear positive relationship with PM 2.5.
Temperature, wind speed, and high solar radiation were found to minimize the spread of COVID-19 cases by Rosario et al. (2020). Air quality, NO2, and PM 2.5 have positive effects on COVID-19 new events, while temperature has negative effects, according to Li et al. (2020).
According to Pei et al. (2021), COVID-19 cases have a clear positive relationship with AQI. Abbreviations: AQI, Air Quality Index; ARDL, Auto Regressive Distributed Lag COVID-19 cases have also clear positive relationship with PM 10, and PM 2.5, as well as a negative relationship with temperature, carbon monoxide, and COVID-19.
Bangladesh's situation is extremely dangerous and disturbing, and the number of infected humans is growing by the day. The number of deaths has been rising in recent days, causing concern among the public. In order to administer and take useful connotations, it is important to predict the pattern of COVID-19 transmission. Unfortunately, there is still a lack of research into the relationship between meteorological influences and COVID-19. The results of some research findings on the relationship between meteorological factors and COVID-19 were mixed.
Furthermore, research findings revealed that the relationship between climatic variables and COVID-19 cases is time and area dependent.
This study used correlation analysis to analyze the strength and direction of the relationship between climatic variables such as temperature, wind speed, humidity, and PM 2.5 in order to arrive at new conclusions and recommendations. Pearson correlation analysis, Spearman correlation analysis, and Kendall's correlation analysis were all used in this phase. The ARDL model was used to look at the lagged impact of climatic variables on COVID-19 in conjunction with certain non-climatic variables.

Overview of Bangladesh
Bangladesh is a beautiful country in South Asia with a middle-income economy. Bangladesh is the eighth most populous nation in the world. Bangladesh has a very high population density as compared to other developing countries. Bangladesh's population is projected to be 166,078,786 Page 10 people as of May 8, 2021 (United Nation data). Middle-aged people account for 164,689,383 people out of the total population. Bangladesh's population accounts for 2.11 percent of the world's total population. Bangladesh has a land area of 56980 square miles. Bangladesh has a population density of 3277 people per square mile and a median age of 27.6 years. Dhaka is Bangladesh's largest city and capital in terms of national, political, economic, and cultural importance.
Bangladesh has a tropical monsoon climate. The overall climate is divided into three distinct seasons: hot and summer months last from March to June, rainy months last from June to October, and winter months last from October to March. Bangladesh's average temperature ranges from 10°C to 40°C, with average humidity ranging from 30% to 80% and average wind speeds ranging from 7km/h to 24km/h.

Data description
The time series data used in this analysis ranged from May 1, 2020 to April 14

Methodology
Three-correlation analysis, such as Pearson correlation, Spearman correlation, and Kendall's rank correlation, were used in the empirical study. The results of the normality check revealed that the data in our study is not normally distributed (Dogan et al., 2020). As a result, three forms of correlation analysis may describe all circumstances.
Page 12 Pearson correlation is the greatest way for measuring the relationship between two variables because it checks for a linear association between them. The Pearson correlation coefficient is the covariance of two variables divided by multiplication of their standard deviation. As a result, Pearson correlation is based on calculating covariance. The correlation coefficient's range is -1 to +1, and it can be calculated as follows: The non-parametric Spearman correlation evaluates the monotonic statistical dependence of two variables. If the correlations between the variables are linear, the Pearson and Spearman correlation coefficients are equivalent. The spearman correlation coefficient has a range of -1 to +1. If the dependent variable grows as the independent variables rise, the Spearman correlation coefficient is positive. If the dependent variable drops as the independent variable grows, the Spearman correlation coefficient is negative. When variables are not linearly connected, the Spearman correlation coefficient is zero. The Spearman correlation coefficient is calculated using the formula below: Where, the total number of observations is indicated by the number n.; = rank ofrank of .
Kendall's rank correlation coefficient is another non-parametric correlation coefficient. It can be used to check the ordinal relationship between variables and analyze the similarity of the Page 13 orderings. Kendall's correlation coefficient is referred to as' tau '( ), and its range is from -1 to +1.
This is how the Kendall's coefficient is defined: The number of pairs is denoted by n.
Finally, the ARDL model was used to examine the lag effect of climatic and non-climatic factors on COVID-19 in this study. COVID-19 infection and transmission requires 2 to 15 days, according to Wu et al. (2020), Wang et al. (2020), and Ma at al. (2020). As a result, the ARDL model was used to examine the delayed effects in this study. We employed four climatic variables including average temperature, average humidity, PM 2.5 averages, as well as average wind speed, as well as two non-climatic variables, population density and human development index as a control variable. The resilience of climatic and non-climatic elements was tested using the individual ARDL equation. The ARDL equation is formulated as follows when population density is a control variable: Where, COVID stands for COVID-19 daily cases, temp stands for average temperature, hum stands for average humidity, wind stands for average wind speed, and pd is for population density.
The ARDL equation looks like this when the human development index is used as a control variable: Where, COVID indicates COVID-19 daily cases, temp indicates average temperature, hum indicates average humidity, wind indicates average wind speed and human indicates human development index.
The ARDL model was used to investigate the lagged impacts of meteorological factors and COVID-19 broad spread. As a result, the following is our fundamental ARDL equation: Where, COVID indicates COVID-19 daily cases, temp indicates average temperature, hum indicates average humidity and wind indicates average wind speed.
Two methods are used to calculate the ARDL equation (6)  daily new confirmed cases. The daily fresh confirmed cases of COVID-19 are plotted against climatic conditions in Figure 1. Climate conditions and COVID-19 prevalence are highly connected, as seen in Figure 3.

Daily adaptation of COVID-19 and climatic variables
The descriptive statistics of climatic parameters and COVID-19 regular cases are presented in COVID-19's Skewness is 1.615, which is not between -0.5 and 0.5, according to the calculations.
It implies that the COVID-19 data is not symmetrical, and that Kurtosis (6.56) is more than 3,

Page 16
indicating that the distribution is wider than the normal distribution. All environmental parameters (excluding temperature) have skewness values between -0.5 and 0.5. The data for all climatic components are symmetrical, and the kurtosis for all climatic factors is less than three, indicating a narrower distribution than the normal distribution, according to the findings.   The study has found a negative, weak and significant association between COVID-19 and PM 2.5 (r=-0.489; p < 0.01).

Correlation between climatic factors and COVID-19
The results of Kendall's rank correlation test are presented in Table 2  In general, the results of three different types of correlation tests are consistent.

ARDL method result and COVID-19
The basic need for time series data to study the order of integration of variables is the unit root test, often known as the stationary test. The Augmented Dickey-Fuller (ADF) and Phillips-Perron (PP) tests were used to conduct our empirical research.  (4), and the final specification of the ARDL model was determined using a general to specific methodology. Furthermore, for optimal lag structure, the Akaike Information Criterion (1981) was adopted.
The results of the symmetric ARDL model for equation (4), which included one non-climatic element, population density, are shown in In addition, the model's stability was tested using the CUSUM and CUSUM Square tests. The results of these experiments are shown in Figure 2, confirming that the model is extremely stable.  Table 7's lower panel.
The CUSUM and CUSUM Square tests are shown in Figure 3 to assess the model's stability and indicate that it is highly stable. covariates. COVID-19 verified instances will result in a 0.59 percent to 1% rise in the daily lag value 1, and COVID-19 confirmed cases will contribute significantly to 0.12 percent, 0.38 percent, 0.14 percent to 1% increases in the daily lag values 6, 7, and 8, respectively, of auto-lag covariate, according to 6a's findings. COVID-19 confirmed cases, on the other hand, will contribute to a drop in the current number of cases by increasing the daily lag values of 9 and 12 by 0.14 percent and 0.122 percent to 1 percent, respectively. Temperature, PM 2.5, and wind speed all rises by 1% in COVID-19 new instances, with increases of 2.40 percent, 0.98 percent, and 1.20 percent, respectively. Furthermore, residual analysis was used to test the ARDL model's reliability.  Figure 4a to assess the model's stability and indicate that it is extremely stable.
COVID-19 confirmed cases would lead to an increase in the current number of cases by 0.58 percent, 0.13 percent, 0.39 percent, and 0.14 percent to 1 percent in the daily lag values of 1, 6, 7, and 8, respectively, for 12 days lag covariates, according to 6b findings. Due to a 12-day covariate lag, COVID-19 confirmed cases would contribute to a 0.12% decline in the current number of cases. Temperature, PM 2.5, and wind speed all rose by 1% in COVID-19 new instances, with increases of 2.59 percent, 0.89 percent, and 1.14 percent, respectively. The ARDL model's dependability was also evaluated using residual analysis. Table 5's lower column contains the results of all testing. Figure 4b shows the CUSUM and CUSUM Square tests, which check the model's stability and confirm that it is highly stable.
Page 21

Discussion
We may conclude from non-meteorological measures that health-policy initiatives that work, protection programs, standard of education and understanding, and population age all play a role in avoiding pandemics and disease crises. The authors recommend that, in order to reduce COVID-19 transmission, the Bangladesh should change policies to take into account Protecting one's health, ensuring one's safety, educating oneself, and aging the population are all factors to consider.
The three statistical correlations for Bangladesh have revealed a substantial relationship between the variables COVID-19, temperature, PM 2.5, wind speed, and humidity. COVID-19 is, in reality, a virus have a strong positive relationship with the temperature, wind speed, and humidity and negative association with PM 2.5. The fact that all of these weather variables have the potential to affect the number of COVID-19 events reported can help to explain this result.
The findings could back up the idea that COVID-19 infection is linked to air pollution.
According to the findings of the ARDL, temperature, wind speed, and PM 2.5 influenced the daily number of COVID-19 verified cases. Indeed, at various degrees of statistical significance, COVID-19 is affected by temperature (or PM 2.5) in a positive (or negative) way (1 percent).
The delayed findings, in particular, back up previous research findings that imply a time lag between previous infections, temperature, wind speed, PM 2.5, and COVID-19 daily-confirmed cases. The incubation phase may be seen as required while developing travel policies and regulations (Xie and Zhu 2020). As a result, those countries that used quarantine and smart lockdown methods to govern movement were able to be successful at managing the COVID-19 situation. These discoveries can be viewed as a contribution to the field of study and they may Page 22 help Bangladesh develop new disease control implications. As a result, intoxication and the weather may be considered COVID-19 defining criteria.
Since South Asian countries have low incomes and a dense population, and people are less aware of the transmission, on sunny days, people crack the lockdown for safety reasons. Moreover, it's possible that one explanation is that there's a connection between temperature and the number of cases registered. One of the most significant indirect factors for COVID-19 transmission in this case is temperature. According to some recent studies, Maximum temperature was found to have a COVID-19 transmission is significantly affected (Shao et al., 2021;Hridoy et al., 2021;Singh et al., 2020;Islam et al., 2020;S.K. Pani et al., 2020;S.Kumar 2020;Xie and Zhu 2020).
Unlike some recent studies that found a negative relationship between temperature and the ability to transmit of COVID-19, this discovery was made (Dogan et al., 2020;Sahoo et al., 2020;Shahzad et al., 2020).
The effect of moisture in the air can be blamed for the overall effects. Exhaled bio-aerosols evaporate quickly in low-humidity conditions, forming nuclei of droplets that can remain in the air and on the ground for extended periods. These nuclei are referred to as persistent nuclei. As a result, virus transmission can be accelerated. Respiratory droplets can pass SARS-CoV-1 from person to person. Because its genetics are similar to SARS-CoV-2, it is thought to be capable of infecting droplets in the lungs. SARS-CoV-2 is a virus that can live for a long time. On plastic, stainless steel, iron, cardboard, and glass, it can last for 3 hours in aerosol form (5 m) and 72 hours in droplet form (> 5 m) according to the World Health Organization (WHO).

Page 23
The occurrence of COVID-19 has no significant association with relative humidity, according to . Other studies have observed the similar findings (Hridoy et al., 2021;Islam et al., 2020;Li et al., 2020;Sajadi et al., 2020;Runkle et al., 2020;Wang et al., 2020;), suggesting that COVID-19 is linked to relative humidity. Dbouk and Drikakis (2020) demonstrated the role of wind speed in SARS-CoV-2 airborne transmission using a computational fluid dynamics simulation. At wind speeds ranging from 4 to 15 km/h, respiratory droplets can travel up to 6 meters, implying that a social distance of 2 meters may not be sufficient. The rising number of incidences of high temperatures and humidity in the southern United States, Brazil, India, and Bangladesh, however, refutes the impact of these conditions.
Maximum wind speed was found to have a major effect on the transmission of COVID-19; this is in line with preceding findings (Hridoy et al., 2021;Ahmadi et al., 2020;Kulkarni et al., 2021;Singh et al., 2020;Islam et al., 2020;Sahin M. 2020). Unlike some recent studies that found no relationship between wind speed and the ability to transmit of COVID-19, this discovery was made (Li et al., 2020), while a negative association was found by Islam et al. (2020).
Furthermore, during previous influenza virus outbreaks or extreme acute respiratory syndrome, the wind speed was taken into consideration as one of the primary factors that aided SARS,

MERS, and influenza transmission (SARS). The probability of COVID-19 virus transmission
can be increased in confined areas with high wind speeds because infectious droplets have a much larger particle density.

Page 24
Despite its many benefits, this research is not without flaws. We've discussed the importance of wind speed, but determining it can be challenging. It's also possible that the direction of the wind plays a role, but we can't investigate its impact on virus transmission due to a lack of data.
Furthermore, our findings are focused on data from the outdoor weather system, which is one of our study's major limitations. Indoor conditions, on the other hand, may have a profound impact on SARS-CoV-2 transmission. When evaluating the relationship between meteorological variables and COVID-19 in Bangladesh, future studies should consider these factors.

Conclusion
The