Hurricane data analyis

Achim Rumberger 27 Januar 2018

Analysis of Hurricane data from HURDAT package

I would like to express my thanks to the contributor of the HURDAT package ("https://rdrr.io/cran/HURDAT/"). Compared to the original data from https://www.nhc.noaa.gov/data/hurdat/ it got so much easier to work with the data.

#libraries
library(tidyverse)

## ── Attaching packages ──────────────────────────────────────────────────────────────── tidyverse 1.2.1 ──

## ✔ ggplot2 2.2.1     ✔ purrr   0.2.4
## ✔ tibble  1.3.4     ✔ dplyr   0.7.4
## ✔ tidyr   0.7.2     ✔ stringr 1.2.0
## ✔ readr   1.1.1     ✔ forcats 0.2.0

## ── Conflicts ─────────────────────────────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()

library(ggthemes)
library(ggmap)
library(htmlwidgets)
library(gridExtra)

## 
## Attaching package: 'gridExtra'

## The following object is masked from 'package:dplyr':
## 
##     combine

library(HURDAT)
library(lubridate)

## 
## Attaching package: 'lubridate'

## The following object is masked from 'package:base':
## 
##     date

library(splines)
library(plotly)

## 
## Attaching package: 'plotly'

## The following object is masked from 'package:ggmap':
## 
##     wind

## The following object is masked from 'package:ggplot2':
## 
##     last_plot

## The following object is masked from 'package:stats':
## 
##     filter

## The following object is masked from 'package:graphics':
## 
##     layout

source("classifiy_hurricianes.R")

Read Inital and Save Data

The data span the time from 1851 to 2016. We will have almost 80k observations with 24 variables. It will take a while to read the inital data with get_hurdat function. Therefore I will save the data for more easy access in the future. If you run this code the first time please uncomment the lines with "get_hurdat" and "saveRDS".

#data <- get_hurdat(basin = c("AL", "EP"))
#saveRDS(data, "data/hurricanesALEP")
data <- readRDS("data/hurricanesALEP")

Manage the dates

data$YEAR <- year(data$DateTime)
data$MONTH <- month(data$DateTime)
data$DAY <- day(data$DateTime)
summary(data %>% select(YEAR, MONTH, DAY, Wind, Pressure))

##       YEAR          MONTH             DAY             Wind       
##  Min.   :1851   Min.   : 1.000   Min.   : 1.00   Min.   : 10.00  
##  1st Qu.:1939   1st Qu.: 8.000   1st Qu.: 8.00   1st Qu.: 30.00  
##  Median :1975   Median : 9.000   Median :16.00   Median : 45.00  
##  Mean   :1964   Mean   : 8.517   Mean   :15.86   Mean   : 51.57  
##  3rd Qu.:1997   3rd Qu.: 9.000   3rd Qu.:23.00   3rd Qu.: 65.00  
##  Max.   :2016   Max.   :12.000   Max.   :31.00   Max.   :185.00  
##                                                  NA's   :338     
##     Pressure     
##  Min.   : 872.0  
##  1st Qu.: 986.0  
##  Median :1000.0  
##  Mean   : 993.5  
##  3rd Qu.:1007.0  
##  Max.   :1024.0  
##  NA's   :43473

First plot to get an overview of the data

# first plot
df = data %>%
  group_by(YEAR) %>%
  summarise(Distinct_Storms = n_distinct(Key))

p = ggplot(df, aes(x = YEAR, y = Distinct_Storms)) + theme_stata()
p + geom_line(size = 1.1) + 
  ggtitle("Number of Storms Per Year") + 
  theme(plot.title = element_text(hjust = 0.5)) +
  geom_smooth(method='loess', se = FALSE, formula = y ~ x) + 
  ylab("Storms")

Categorize storms

Create a new variable which classifies the hurricanes according to windspeed.

data$CATEGORY <- sapply(data$Wind, classifyHurricane)

Plot the storms by category

#by category
df = data %>%
  filter(grepl("H", CATEGORY)) %>%
  group_by(YEAR,CATEGORY) %>%
  summarise(Distinct_Storms = n_distinct(Key))
df$CATEGORY = factor(df$CATEGORY)

p = ggplot(df, aes(x = YEAR, y = Distinct_Storms, col = CATEGORY)) + theme_stata()
p + geom_line(size = 1.1) + 
  scale_color_brewer(direction = -1, palette = "Dark2") + 
  ggtitle("Number of Storms Per Year By Category (H)") + 
  theme(plot.title = element_text(hjust = 0.5)) +
  facet_wrap(~ CATEGORY, scales = "free_x", ncol = 1) + 
  geom_smooth(method = 'lm', se = FALSE, col = 'black') +
  theme(axis.text.x = element_text(hjust = 1, angle=45), legend.position = 'none') + 
  ylab('Number of Storms') +
  theme(axis.title.y = element_text(margin = margin(t = 0, r = 10, b = 0, l = 0)))

## Make a distinct dataset I also will filter the data. A hurricane in its lifecycle will go through several stages, from tropical storm to a class H1 hurricane to H2 to H1, tropical storm. I want to count each storm only once with its highest ratings. On my laptop this takes a while, so again, I will save the data for further analysis. This will show us, that in the time frame from 1851 to 2016 a total of 2900 storms have been recorded. If you run this code the first time please uncomment the lines with "highestClassification4Storm" and "saveRDS".

#distinctStormData <- highestClassification4Storm(data)
#saveRDS(distinctStormData, "data/hurricanesALEPAAA")
distinctStormData <- readRDS("data/hurricanesALEPAAA")
summary(distinctStormData)

##      Key                 YEAR        CATEGORY              Wind       
##  Length:2900        Min.   :1851   Length:2900        Min.   :  0.00  
##  Class :character   1st Qu.:1942   Class :character   1st Qu.: 45.00  
##  Mode  :character   Median :1974   Mode  :character   Median : 60.00  
##                     Mean   :1963                      Mean   : 67.45  
##                     3rd Qu.:1996                      3rd Qu.: 90.00  
##                     Max.   :2016                      Max.   :185.00

Plot all distinct storms

df = distinctStormData %>%
  group_by(YEAR) %>%
  summarise(Distinct_Storms = n_distinct(Key))

p = ggplot(df, aes(x = YEAR, y = Distinct_Storms)) + theme_stata()
p + geom_line(size = 1.1) + 
  ggtitle("Number of distinct Storms Per Year") + 
  theme(plot.title = element_text(hjust = 0.5)) +
  geom_smooth(method='loess', se = FALSE, formula = y ~ x) + 
  ylab('Number of Storms') +
  theme(axis.title.y = element_text(margin = margin(t = 0, r = 10, b = 0, l = 0)))

## Plot disitinct hurricanes by category

df = distinctStormData %>%
  filter(grepl("H", CATEGORY)) %>%
  group_by(YEAR,CATEGORY) %>%
  summarise(Distinct_Storms = n_distinct(Key))
df$CATEGORY = factor(df$CATEGORY)

p = ggplot(df, aes(x = YEAR, y = Distinct_Storms, col = CATEGORY)) + theme_stata()
p + geom_line(size = 1.1) + 
  scale_color_brewer(direction = -1, palette = "Dark2") + 
  ggtitle("Number ofd disitinct Storms Per Year By Category (H)") + 
  theme(plot.title = element_text(hjust = 0.5)) +
  facet_wrap(~CATEGORY, scales = "free_x", ncol = 1) + 
  geom_smooth(method = 'lm', se = FALSE, col = 'black') +
  theme(axis.text.x = element_text(angle=45), legend.position = 'none') + 
  ylab('Number of Storms') +
  theme(axis.title.y = element_text(margin = margin(t = 0, r = 10, b = 0, l = 0)))

Discussion

• Noteworthy to me seem to be two observations:
• the marked increase in overall storm activity from the 1950s
• the increase in category 5 hurricanes. What was once a very event, from the 1990s onward the occurence of a major hurricane became a regular feature of the hurricane season.