As part of this assignment, I needed to produce an informative graphical visualization that reflected the relationship between any two variables in the context of Earth science. To increase the relevance of the assignment, it was decided to focus on a metric related to human life, namely mortality. Mortality is commonly referred to as the number of human deaths per 1,000 lives, so this variable tends to be measured in small values. Mortality itself reflects increased or decreased mortality trends and is a polysemous term explaining why. The 72 years of data came from the Macrotrends website, which cited the UN report on annual global mortality (UN, 2022). At first glance, this data signals that human mortality is slowly falling over time (Appendix A), which may be due to changes in healthcare policy, increased technological equipment in hospitals and increased availability of clinical services. The second variable is annual temperature anomalies averaged over all countries (NASA GISS, 2022). A temperature anomaly is a difference in mean annual temperature between the current year and the previous year for a particular location. For example, if the average annual temperature in the United States in 2021 was 20°C, and in 2022 it was already 22°C, then the temperature anomaly in 2022 was +2°C. It is essential that this value can also be harmful, including if the current year turns out to be, on average colder than the previous year. From Appendix A, we can see that this parameter shows an increasing trend over time, which means that the Earth’s surface is getting hotter.
In this paper, these two variables were motivated primarily by my desire to investigate data that are not obvious to analyze but make me think and dig deeper. For this reason, I chose as the independent variable mean annual temperature anomalies the tendency of the planet to cool or cool dynamically. As a dependent variable, the choice of global mortality was motivated by the desire to determine whether increased or decreased mortality could be a consequence of temperature dynamics. Figure 1 below reflects the constructed visualization and provides information about the regression line with the equation and coefficient of determination.
As can be seen, the graph contains information on multiple points, namely for 72 years, from 1950 to 2021. The overall trend observed is reflected by a linear regression model: an increase in the value of temperature anomalies (warming of the planet) leads to a decrease in mortality. The linear model describes only up to 64% of the variance of these points, and if you pay attention to the right end of the data set, you can see that mortality is reaching some plateau after a rather steep parabolic decline. The relationship between these two factors can thus be postulated to be inversely proportional: an increase in one variable leads to a decrease in the other. In order to be convinced of the strong relationship between the variables, the Pearson correlation coefficient was calculated (Laerd Statistics, 2021). The value did justify a strong relationship and was -0.80, the negative sign of which indicates an inverse relationship between the variables.
Based on probabilistic predictions, regression analysis helps estimate the relationship’s past and future development. Since it is already known that the planet is warming rapidly over time, we should expect future human mortality to be even lower or to plateau more steadily. On the other hand, the planet was colder in the past, and thus the mortality rate turned out to be higher. Much of this assessment is justified by the increased availability of medicine that has been observed over time — more people are receiving clinical services, which reduces mortality. However, it is worth assuming that the planet’s continuing warming could cause severe ecological shifts, including melting glaciers, drought, global famine, pandemics, and the destruction of the ozone layer. Consequently, mortality is supposed to rise again over time after we leave the plateau.
References
Laerd Statistics. (2021). Pearson product-moment correlation. Laerd Statistics.
NASA GISS. (2022). Global temperature. Global Climate Change NASA. Web.
UN. (2022). World population prospects: World death rate 1950-2022. Macrotrends. Web.