Synopsis

This analysis uses the complete NOAA Storm Database supplied for the assignment to identify the event types associated with the greatest population-health and economic impacts across the United States. Population-health impact is measured as the sum of reported fatalities and injuries. Economic impact is measured as the sum of reported property and crop damage after converting the magnitude codes to dollar multipliers. Event labels are standardized only by trimming whitespace and converting them to uppercase, so the results remain tied to the database’s EVTYPE categories. Tornadoes produce the largest combined number of fatalities and injuries. Floods produce the greatest total reported economic damage. The results should be interpreted with care because event naming and reporting practices vary across the historical record.

Data Processing

The analysis begins with the raw compressed CSV file. If it is not already in the working directory, the code downloads the assignment file; read.csv() then reads the CSV directly through an bzfile() connection. Only the fields needed for the two questions are retained. Missing numerical values are replaced by zero, and event names are converted to uppercase after surrounding whitespace is removed.

knitr::opts_chunk$set(
  echo = TRUE,
  warning = FALSE,
  message = FALSE,
  fig.width = 9,
  fig.height = 6
)
options(scipen = 999)
data_file <- "repdata_data_StormData.csv.bz2"
data_url <- paste0(
  "https://d396qusza40orc.cloudfront.net/",
  "repdata%2Fdata%2FStormData.csv.bz2"
)

if (!file.exists(data_file)) {
  download.file(data_url, data_file, mode = "wb")
}

storm_raw <- read.csv(
  bzfile(data_file),
  stringsAsFactors = FALSE,
  check.names = FALSE
)

needed_columns <- c(
  "EVTYPE", "FATALITIES", "INJURIES",
  "PROPDMG", "PROPDMGEXP", "CROPDMG", "CROPDMGEXP"
)
storm <- storm_raw[, needed_columns]
rm(storm_raw)

storm$EVTYPE <- toupper(trimws(storm$EVTYPE))
storm$FATALITIES[is.na(storm$FATALITIES)] <- 0
storm$INJURIES[is.na(storm$INJURIES)] <- 0
storm$PROPDMG[is.na(storm$PROPDMG)] <- 0
storm$CROPDMG[is.na(storm$CROPDMG)] <- 0

Population-health consequences are calculated by adding fatalities and injuries for each record and then summing by event type.

storm$HEALTH_IMPACT <- storm$FATALITIES + storm$INJURIES

health_by_event <- aggregate(
  cbind(FATALITIES, INJURIES, HEALTH_IMPACT) ~ EVTYPE,
  data = storm,
  FUN = sum,
  na.rm = TRUE
)
health_by_event <- health_by_event[
  order(health_by_event$HEALTH_IMPACT, decreasing = TRUE),
]
top_health <- head(health_by_event, 10)
rownames(top_health) <- NULL

The damage exponent fields contain letters and, in a small number of records, digits. The following function converts H, K, M, and B to hundreds, thousands, millions, and billions. Numeric exponent codes are interpreted as powers of ten; blank or unrecognized codes contribute a multiplier of one so that their recorded base damage is not discarded.

damage_multiplier <- function(exponent) {
  code <- toupper(trimws(as.character(exponent)))
  multiplier <- rep(1, length(code))

  multiplier[code == "H"] <- 100
  multiplier[code == "K"] <- 1000
  multiplier[code == "M"] <- 1000000
  multiplier[code == "B"] <- 1000000000

  numeric_code <- grepl("^[0-9]$", code)
  multiplier[numeric_code] <- 10 ^ as.numeric(code[numeric_code])
  multiplier
}

storm$PROPERTY_DOLLARS <- storm$PROPDMG * damage_multiplier(storm$PROPDMGEXP)
storm$CROP_DOLLARS <- storm$CROPDMG * damage_multiplier(storm$CROPDMGEXP)
storm$TOTAL_DOLLARS <- storm$PROPERTY_DOLLARS + storm$CROP_DOLLARS

economic_by_event <- aggregate(
  cbind(PROPERTY_DOLLARS, CROP_DOLLARS, TOTAL_DOLLARS) ~ EVTYPE,
  data = storm,
  FUN = sum,
  na.rm = TRUE
)
economic_by_event <- economic_by_event[
  order(economic_by_event$TOTAL_DOLLARS, decreasing = TRUE),
]
top_economic <- head(economic_by_event, 10)
rownames(top_economic) <- NULL

Results

Events most harmful to population health

The ten event types with the largest combined numbers of fatalities and injuries are shown below. Tornadoes rank first by a wide margin.

knitr::kable(
  top_health,
  format.args = list(big.mark = ","),
  col.names = c("Event type", "Fatalities", "Injuries", "Total health impact"),
  caption = "Ten event types with the greatest population-health impact"
)
Ten event types with the greatest population-health impact
Event type Fatalities Injuries Total health impact
TORNADO 5,633 91,346 96,979
EXCESSIVE HEAT 1,903 6,525 8,428
TSTM WIND 504 6,957 7,461
FLOOD 470 6,789 7,259
LIGHTNING 816 5,230 6,046
HEAT 937 2,100 3,037
FLASH FLOOD 978 1,777 2,755
ICE STORM 89 1,975 2,064
THUNDERSTORM WIND 133 1,488 1,621
WINTER STORM 206 1,321 1,527 
health_plot <- top_health[order(top_health$HEALTH_IMPACT), ]
par(mar = c(5, 12, 3, 2))
barplot(
  health_plot$HEALTH_IMPACT,
  names.arg = health_plot$EVTYPE,
  horiz = TRUE,
  las = 1,
  col = "firebrick3",
  border = NA,
  xlab = "Fatalities plus injuries",
  main = "Population-health impact by event type"
)
Event types with the greatest combined number of fatalities and injuries.
Event types with the greatest combined number of fatalities and injuries.

Events with the greatest economic consequences

The following table reports property damage, crop damage, and their combined total in billions of dollars for the ten highest-impact event types. Floods have the largest total economic impact in this database, followed by hurricanes/typhoons and tornadoes.

economic_table <- top_economic
economic_table[, c("PROPERTY_DOLLARS", "CROP_DOLLARS", "TOTAL_DOLLARS")] <-
  economic_table[, c("PROPERTY_DOLLARS", "CROP_DOLLARS", "TOTAL_DOLLARS")] / 1000000000

knitr::kable(
  economic_table,
  digits = 2,
  format.args = list(big.mark = ","),
  col.names = c(
    "Event type", "Property damage ($ billions)",
    "Crop damage ($ billions)", "Total damage ($ billions)"
  ),
  caption = "Ten event types with the greatest reported economic impact"
)
Ten event types with the greatest reported economic impact
Event type Property damage ($ billions) Crop damage ($ billions) Total damage ($ billions)
FLOOD 144.66 5.66 150.32
HURRICANE/TYPHOON 69.31 2.61 71.91
TORNADO 56.95 0.41 57.36
STORM SURGE 43.32 0.00 43.32
HAIL 15.74 3.03 18.76
FLASH FLOOD 16.82 1.42 18.24
DROUGHT 1.05 13.97 15.02
HURRICANE 11.87 2.74 14.61
RIVER FLOOD 5.12 5.03 10.15
ICE STORM 3.94 5.02 8.97 
economic_plot <- top_economic[order(top_economic$TOTAL_DOLLARS), ]
par(mar = c(5, 12, 3, 2))
barplot(
  economic_plot$TOTAL_DOLLARS / 1000000000,
  names.arg = economic_plot$EVTYPE,
  horiz = TRUE,
  las = 1,
  col = "steelblue3",
  border = NA,
  xlab = "Total damage (billions of U.S. dollars)",
  main = "Economic impact by event type"
)
Event types with the greatest combined property and crop damage.
Event types with the greatest combined property and crop damage.

The ranking uses reported nominal damage values and does not adjust for inflation. It also preserves the event categories found in EVTYPE, apart from capitalization and whitespace, so synonymous or historically inconsistent labels are not combined.