Introduction:

Link to original article:https://www.r-bloggers.com/plotting-the-impact-of-atlantic-hurricanes-on-the-us/

Throughout the years, hurricanes from the Atlantic Ocean continue to make an impact on the United States. This article was written with hurricane Florence right around the corner which sparked interest to analyze past hurricanes to determine if there are certain characteristics of hurricanes that make them so devastating. Historical data dating back to hurricane Betsy in 1965 was used to graph different characteristics like fatalities, damages, wind speed and pressure.

Analysis:

In recent history, hurricanes Katrina and Maria are the top two deadliest storms. Fatality rates for both storms reach approximately 1,800 and 3,000 deaths respectively. Meanwhile many of the other recorded hurricanes have less than 200 deaths.

ggplot(Hurricanes, aes(x=Year, y=US_Fatalities, label=Name)) + 
  geom_point(color = 'red') + 
  geom_text_repel(size=5) + theme_bw() + theme(text = element_text(size=20)) +
  ggtitle("US Fatalities")

Same as above graph, but adjusted to unclutter hurricanes with lower fatalities.

ggplot(Hurricanes, aes(x=Year, y=US_Fatalities, label=Name)) + 
  geom_point(color = 'red') + scale_y_log10(breaks = c(0,1,10,100,1000,2000,3000)) +
  geom_text_repel(size=5) + theme_bw() + theme(text = element_text(size=20)) +
  ggtitle("US Fatalities log10")
## Warning: Transformation introduced infinite values in continuous y-axis

## Warning: Transformation introduced infinite values in continuous y-axis

While Katrina and Maria are the top two deadliest hurricanes recorded, other hurricanes with much lower death tolls rival them in damages. Hurricane Harvey, Sandy and Irma stick out as storms that did not take as many lives but still caused billions in damage.

 ggplot(Hurricanes, aes(x=Year, y=Damage_billions, label=Name)) + 
  geom_point(color = 'red') + 
  geom_text_repel(size=5) + theme_bw() + theme(text = element_text(size=20)) +
  ggtitle("Damages in Billions of USD")

Another way to visualize the death tolls is by a bar graph. All storms with less than 200 deaths were grouped into one column, totaling 40 storms and 1,200 deaths. It highlights how deadly Katrina and Maria truly are since 40 storms do not even add up to the lives taken by either one of the two largest storms. The question that is brought up here is how a storm can cause so much physical damage without also taking a proportionate amount of lives.

Hurricanes %>% filter(!(Name %in% c("Maria", "Katrina"))) %>% summarize(sum(US_Fatalities))
##   sum(US_Fatalities)
## 1               1773
Hurricanes$Hurricane = Hurricanes$Name
Hurricanes[rank(-Hurricanes$US_Fatalities)>5,]$Hurricane = "All other hurricanes (40)"
  ggplot(Hurricanes, aes(Hurricane, US_Fatalities)) + 
  theme_bw() + theme(text = element_text(size=15)) +
  geom_col()  + ggtitle("US Hurricane Deaths")

The last two figures go hand in hand, but are both difficult to analyze. Storms that have statistically insignificant death counts have astronomically high fatality to wind ratios. With the information available to the public, it is difficult to discern exactly what components lead to a deadlier or more destructive storm.

 ggplot(Hurricanes, aes(x=Highest_winds_mph, y=US_Fatalities, label=Name)) + 
  geom_point(color = 'red') + 
  geom_text_repel(size=5) + theme_bw() + theme(text = element_text(size=20)) +
  ggtitle("Highest Winds MPH against US Fatalities")

ggplot(Hurricanes, aes(x=Lowest_Pressure_mbar, y=Fatalities, label=Name)) + 
  geom_point(color = 'red') + 
  geom_text_repel(size=5) + theme_bw() + theme(text = element_text(size=20)) +
  ggtitle("Lowest Pressure (mbar) against Fatalities")

Follow Up:

We decided to make a bar graph to show the damages done by the top five hurricanes. We feel like the bar graph shows a dramatic difference in damage cost among these hurricanes. Throughout the rest of the data we saw that hurricanes Katrina and Maria stood out from the rest. However, when it comes to damage cost there are other hurricanes that stand out with them. The purpose of the article was to find characteristics of hurricanes that caused them to be so devastating. Hurricane Katrina caused a lot of damage and a lot of deaths. Meanwhile hurricanes like Irma did enough damage to get in to this top 5 and had less than 100 deaths. We can conclude that damages is not a reliable characteristic to determine deaths.

Hurricanes$HurricaneD <- Hurricanes[rev(order(Hurricanes$Damage_billions)),"Name"][1:5]
  ggplot(Hurricanes, aes(HurricaneD, Damage_billions)) + 
  theme_bw() + theme(text = element_text(size=15)) +
  geom_col()  + labs(title = "Damages in Billions", x = "Top 5 Hurricanes", y = "Damages in Billions (USD)")

Discussion:

As we mentioned above, hurricane damages is not a reliable characteristic to determine deaths. Also there is not enough data publically available to truly determine whether low pressure or high winds will be a more deadly or costly storm. If we did this again, we would instead pick an article that used more data and actually came to a definitive conclusion. The conclusion of “We don’t really have enough data to make a firm conclusion” is not particularly compelling to read.