Nowell Phelps August 15/16 2024
This repository contains code to scrape the Wikipedia page https://en.wikipedia.org/wiki/List_of_natural_disasters_by_death_toll for the distasters with highest deathtolls for each year in the 20th and 21st centuries, and to plot them.
The plots are shown in the pdf "Disaster plots". Figure 1, on the first page of the pdf, shows the deathtoll due to the most deadly natural disaster for each year. Each bar represents one natural disaster. The height of the bar shows the number of deaths attributed to the disaster, and the position of the bar on the x axis (its horizontal position) shows the year it took place. There is only one bar for each year because we are only looking at the most deadly natural disaster in each year. The colour of the bar shows what type of disaster it was. For example, the colour black represents earthquakes and purple represents tropical cyclones. All of the colours and types are explained in the legend on the right hand side of the chart. In some cases, a disaster is a combination of multiple types, such as a tsunami and an earthquake.
We can note that by far the most costly disaster, in terms of human lives lost was in 1931. This is the tall blue bar which extends far higher than all of the others. This bar represents the 1931 China floods, which were estimated to cost between 422,499 and 4 million lives in China. All other disasters took less than 500,000 lives (represented by the grey dashed horizontal line). Because this disaster had a much higher death-toll than the rest, it makes it difficult to see the types of the other disasters and the number of deaths they caused.
Figure 2 (on the second page of the PDF) is the same as Figure 1, but with the y axis capped at 500,000. This means that it doesn't accurately represent the toll of deaths in the 1931 disaster (which again is the blue bar extending to the top of the chart) but it is accurate for all the other disasters and shows their death-toll and types better than Figure 1. It is important when reading this graph to remember that it doesn't show the full extent of the deaths in 1931!
Note that in this analysis epidemics and famine are not considered to be natural disasters and so are not included. Also note that in some cases researchers aren't sure exactly what the number of deaths was. If the number of deaths was not exactly known but there was an estimated range of deaths (like for the 1931 China floods), we used the middle of the range for plotting. If there was an open-ended estimate (for example, 20,000+ people), we used the lower end of the range (20,000).