Effect of Air Pollution on the Environment: Determining the Correlation between Surface Temperature and Carbon Dioxide Concentration

Introduction

Air pollution contributes to climate change, and various components of these pollutants correspond to a significant rise to temperature change in the environment. Among air pollutants, several gases, including particulate matter (PM) and greenhouse gases, contribute largely to the global warming effect. Global warming and climate change are terms used synonymously to refer to the phenomenon by which land and water surface heat gets trapped within the Earth’s atmosphere by a sheet of accumulated greenhouse gases and PM. The net effect is a rise in temperature since the confined gases provide an insulating and filtering mechanism against heat rays from the sun. This is because the gases allow sunlight to pass into the atmosphere but prevent them from leaving it.

Greenhouse gases constitute water vapor, carbon dioxide, methane, ozone gases, nitrous oxide, and chlorofluorocarbons (US EPA, OA, 2018). Carbon dioxide concentration within the atmosphere shows the strongest correlational relationship to global temperature spikes. The National Centers for Environmental Information observe that there is a ‘strong correspondence between temperature and the concentration of carbon dioxide in the atmosphere’ (NCEI, n.d.). Additionally, as of 2018, carbon dioxide accounted for approximately 81% of all pollutant emissions (US EPA, OA, 2018). Therefore, for this research, the two variables of interest will be the global carbon dioxide levels and the combined land-surface air and sea-surface water temperatures spanning 30 years (from 1985 to 2014).

Several studies have looked at how increasing levels of carbon dioxide and temperature have affected the environment. In Soares’ research article, ‘Warming Power of CO₂ and H₂O: Correlations with Temperature Changes’, he denies the verifiability of a relationship between carbon dioxide emission and temperature rise but maintains that the two variables are closely related (Soares, 2010). The study finds that water has a greater impact on temperature rise than carbon dioxide. Stinziano & Way (2014) implicitly show that carbon dioxide levels and temperature changes are correlated in their study on how the two variables affect the growth of boreal forests. Important to the scope of this research, their study focuses on carbon sequestration, which deals with converting carbon into inactive forms to mitigate the effects of carbon dioxide on global warming. Their study complies with the understanding that carbon emissions directly influence the rate of temperature change on a global scale.

Hypothesis and Scope

The study seeks to identify the relationship between global temperature and atmospheric carbon dioxide concentration levels as a subject of air pollution’s effects on the environment. Based on the findings, the hypothesis in question states that there is a positive correlation between the amount of atmospheric carbon dioxide and temperature. The results will determine whether carbon dioxide affects the environment in terms of atmospheric heat, depending on the hypothesis’s validity. Some of the reasons why this hypothesis is thought to be valid are:

1. Greenhouse gases are the main cause behind climate change
2. The global temperature means over the past 20 centuries has been increasing hand in hand with the concentration of greenhouse gases prevalent in the atmosphere
3. Carbon dioxide is the largest contributor to air pollutants among the greenhouse gases
4. The recent spike in carbon dioxide emissions at the turn of the century corresponds to the elated rate of temperature rise (Wheeling Jesuit University, 2014).
5. Some greenhouse gases show steady growth curves and slower growth rates than carbon dioxide, an example being methane (Wheeling Jesuit University, 2014).

Carbon dioxide is the independent variable (x), and the temperature is the dependent variable (y). The measurement of atmospheric temperatures is limited to a global scope. There exist two other regions that shall not be the focus of this study; Northern and Southern Hemisphere temperature means. Extraneous variables include other pollutants that directly affect the increasing global temperatures, and this includes the greenhouse gases and particulate matter. The data in use is from two separate datasets; one for the combined land-surface air and sea-surface water temperatures and the other for global carbon dioxide levels. Datasets are provided by the NASA Goddard Institute for Space Studies (2010) and the CO₂ Earth (2020) foundation, respectively. Given the time-interval nature of the study and datasets, the sampling method used is systematic. Global temperature levels, measured in Degrees Celsius (⁰C), were sampled at distinct monthly intervals, which were then used to compute the annual temperature means. Similarly, carbon dioxide concentration levels were measured in atmospheric parts per million (ppm) on a systematic sampling structure in annual intervals.

 Analysis

One-Variable Analysis

Global Temperatures (1985 – 2014)

Table 1: Global average land-sea temperatures

Mean: 0.48 ⁰C. The value is a measure of the central point from which other temperature values are distributed. Median: 0.50 ⁰C. In 30 years, all recorded global temperatures are equally divided between 0.50 ⁰C; they are greater than or less than the stated value. Mode: 0.32 ⁰C is the most recurrent recorded value, both in the years 1987 and 1994.Standard Deviation: 0.1784. The standard deviation is greater than 1/3 of the mean, implying that the distribution of global temperatures for the 30 years was widely spread away from the mean to a fair degree. Variance: 0.0307.Q1, Q3: 0.345, 0.6375. These represent the lower and upper quartile values, as shown in the box plot.

As depicted in the box plot, the variance of values from the mean in the 2^nd and 3^rd quartiles are fairly equal; there being a greater variance in the 2^nd quartile, which means that global temperatures that are greater than the mean have less variance than those that are lower than the mean. Additionally, the plot shows that the overall spread of values about the mean is low.

The histogram shows an asymmetric distribution curve that is negatively skewed. Pearson’s 2^nd Coefficient of Skewness was used to determine the graph’s skewness;

Skewness = 3 * (Mean – Median) / Standard Deviation

When calculated, the formula gives a skewness of -0.3577, which confirms that the distribution curve is negatively skewed. Since skewness is used to perform base predictions (“Learn about Skewness,” 2019), the right tapering edge of the curve can be used to indicate that Global Temperatures are steadily rising with time.

Carbon Dioxide Concentration (1985 – 2014)

Table 2: Global Carbon Dioxide Concentration

Mean: Carbon dioxide levels are central to the value of 368.836 ppm. Median: 367.8450 ppm represents the equidistant point from which other values are either greater or smaller. Mode: There is no recurrent carbon dioxide concentration recorded over the 30 years. Standard Deviation: 16.0458. Variance: 249.1656. The standard deviation and variance for carbon levels are high, indicating a high spread of values from the mean. Q1, Q3: 355.7909, 381.8065. These represent the lower and upper quartile values, as shown in the box plot.

Like the temperature box plot, the variance of values in the upper and lower quartiles is fairly consistent. Similarly, the spread of values about the mean is fairly high. The histogram, unlike its temperature counterpart, indicates a positive skewness of 0.2042. Such a distribution curve (positive skewness) indicates that a greater increase in the mean will bring about a greater increase in hot extremes (Sura, 2012).

Two-Variable Analysis

Scatter Plot:

As depicted by the trend line on the scatter plot, the variables temperature and carbon dioxide concentration are positively correlated (Cheusheva, 2018). A positive correlation means that as carbon dioxide concentration increases, so do the surface temperatures. The equation of the trend line,   y = 0.0101x – 3.2463, can be used to explain the relationship between the two variables (Cheusheva, 2018). For trends with a near-perfect correlation, the equation of the trend line can be used to predict future values accurately. In the case of the variables plotted above, the y and x values represent the relationship between the independent and dependent variables to the linear regression model equation. For the R Square, otherwise termed the coefficient of determination, the value represents the percentage of points in symmetry with the trend line. An R Square of 78%, indicates a reasonable number of points lie close to/on the trend line.

Correlation

Using Excel’s Correlation function add-in (Data Analysis Tool), a correlation between the variables in question was shown to be 0.8839. The strongest form of positive correlation (perfect correlation) between two variables is 1; hence 0.8839 shows a strong correlation between atmospheric carbon dioxide and surface temperature.

It can be proven that a relationship between the amount of surface temperature and carbon dioxide exists, where the amount of carbon dioxide concentration positively correlates to atmospheric temperature. Further, the linear regression equation can be used to extrapolate the expected temperature means at given carbon dioxide concentrations. For example, at concentrations of 410, temperature levels can be expected to rise by 0.8 ⁰C.

Conclusion

Through exploratory analysis, the study identified a strong correlation between the levels of global atmospheric carbon dioxide levels and surface temperatures, thereby justifying the hypothesis that there is a positive correlation between the amount of atmospheric carbon dioxide and temperature. The study involved both one-variable and two-variable analyses in which the trends and measures of central tendency for the variables were computed in the one-variable analysis phase. Similar characteristics for both temperature and carbon dioxide were seen in both one-variable analyses with measures such as variance and mean similarity, as depicted by the box plots. For the two-variable analyses, Excel’s correlation function proved a strong mutual relationship between the two variables.

This correlation was backed up by the trend line formed by the scatter plot that showed a positive linear correlation, which implies that as carbon dioxide increases, so does the atmospheric temperature. Concerning the subject of the study, and having tested and justified the hypothesis, it is equally justifiable to state that carbon dioxide as an air-polluting substance negatively affects the environment. An increase in its atmospheric concentration corresponds to an increase in atmospheric temperature. Bias was prevalent in the reliance of data from multiple sources of information for the datasets, whose figures might not be accurately presented. Future studies in the same research field should use multiple resources for the same data source to improve reliability. It is recommended that a study be carried out to determine whether the relationship between carbon dioxide levels and surface temperatures have a causal relationship to further this research.

References

Cheusheva, S. (2018, October 3). How to create a scatter plot in Excel. Retrieved July 30, 2020, from Excel tutorials, functions and formulas for beginners and advanced users – Ablebits.com Blog website: https://www.ablebits.com/office-addins-blog/2018/10/03/make-scatter-plot-excel/

CO2 Earth. (2020). Historical CO2 Datasets. Retrieved July 30, 2020, from CO2.Earth website: https://www.co2.earth/historical-co2-datasets

Goddard Institute for Space Studies. (2010). Data. GISS:  GISS Surface Temperature Analysis, GISTEMP/v3. Retrieved from Nasa.gov website: https://data.giss.nasa.gov/gistemp/

Learn About Skewness. (2019). Retrieved from Investopedia website: https://www.investopedia.com/terms/s/skewness.asp

National Centers for Environmental Information (NCEI). (n.d.). Temperature Change and Carbon Dioxide Change. Retrieved July 30, 2020, from www.ncdc.noaa.gov website: https://www.ncdc.noaa.gov/global-warming/temperature-change#:~:text=When%20the%20carbon%20dioxide%20concentration%20goes%20up%2C%20temperature%20goes%20up.&text=A%20small%20part%20of%20the

Soares, P. C. (2010). Warming Power of CO2 and H2O: Correlations with Temperature Changes. International Journal of Geosciences, 01(03), 102–112. https://doi.org/10.4236/ijg.2010.13014

Stinziano, J. R., & Way, D. A. (2014). Combined effects of rising CO2 and temperature on boreal forests: growth, physiology, and limitations. Botany, 92(6), 425–436. https://doi.org/10.1139/cjb-2013-0314

Sura, S. (2012). Extreme temperatures in a warming climate: modulation of the effect of changes in the mean and variance by the distribution skewness.

US EPA, OA. (2018, October 31). Overview of Greenhouse Gases | US EPA. Retrieved from US EPA website: https://www.epa.gov/ghgemissions/overview-greenhouse-gases

Wheeling Jesuit University. (2014). Greenhouse Gases and Temperature. Retrieved from Cet.edu website: http://ete.cet.edu/gcc/?/globaltemp_ghgandtemp/

‌