Synopsis

This report analyzes the U.S. National Oceanic and Atmospheric Administration (NOAA) Storm Database to identify which types of severe weather events are most harmful to population health and which have the greatest economic consequences. The raw data span 1950 to November 2011 and record fatalities, injuries, and property/crop damage estimates by event type (EVTYPE). After loading the raw compressed CSV file directly into R, the event-type field was standardized, and property/crop damage values were converted to dollar amounts using their associated magnitude codes (K, M, B, etc.). Health impact was assessed by summing fatalities and injuries per event type, and economic impact was assessed by summing property and crop damage per event type. The results show that tornadoes are by far the leading cause of both fatalities and injuries, while floods cause the greatest total economic damage, with hurricanes/typhoons and tornadoes also ranking among the costliest event types. These findings can help government and municipal managers prioritize resources for severe weather preparedness.

Data Processing

Loading the raw data

The analysis starts directly from the raw, compressed CSV file provided by the course. If the file is not already present in the working directory, it is downloaded automatically. The file is read directly from its .bz2 compressed form using bzfile(), with no manual/external preprocessing.

library(dplyr)
library(ggplot2)
library(gridExtra)
library(knitr)
library(scales)

dataFile <- "StormData.csv.bz2"

if (!file.exists(dataFile)) {
    download.file(
        "https://d396qusza40orc.cloudfront.net/repdata%2Fdata%2FStormData.csv.bz2",
        destfile = dataFile, method = "libcurl", mode = "wb"
    )
}

storm <- read.csv(bzfile(dataFile), stringsAsFactors = FALSE)
dim(storm)
## [1] 902297     37
str(storm[, c("BGN_DATE","EVTYPE","FATALITIES","INJURIES","PROPDMG","PROPDMGEXP","CROPDMG","CROPDMGEXP")])
## 'data.frame':    902297 obs. of  8 variables:
##  $ BGN_DATE  : chr  "4/18/1950 0:00:00" "4/18/1950 0:00:00" "2/20/1951 0:00:00" "6/8/1951 0:00:00" ...
##  $ EVTYPE    : chr  "TORNADO" "TORNADO" "TORNADO" "TORNADO" ...
##  $ FATALITIES: num  0 0 0 0 0 0 0 0 1 0 ...
##  $ INJURIES  : num  15 0 2 2 2 6 1 0 14 0 ...
##  $ PROPDMG   : num  25 2.5 25 2.5 2.5 2.5 2.5 2.5 25 25 ...
##  $ PROPDMGEXP: chr  "K" "K" "K" "K" ...
##  $ CROPDMG   : num  0 0 0 0 0 0 0 0 0 0 ...
##  $ CROPDMGEXP: chr  "" "" "" "" ...

Selecting relevant variables

For this analysis we only need the event type and the health/economic impact variables, so we subset the data to reduce memory overhead.

storm2 <- storm %>%
    select(EVTYPE, FATALITIES, INJURIES, PROPDMG, PROPDMGEXP, CROPDMG, CROPDMGEXP)

Cleaning the EVTYPE field

The EVTYPE field contains inconsistent capitalization and extra whitespace (e.g. “TSTM WIND”, “tstm wind”, “Thunderstorm Wind”), which would otherwise fragment identical event types into separate categories. We standardize case and trim whitespace. A full free-text remapping of all ~900 raw EVTYPE values to the 48 official categories is beyond the scope of this report, but basic normalization substantially reduces duplication and is sufficient to identify the top contributing event types, since the highest-impact categories are already recorded fairly consistently in the raw data.

storm2$EVTYPE <- trimws(toupper(storm2$EVTYPE))

Converting property and crop damage to dollar values

PROPDMG/CROPDMG give a numeric magnitude, and PROPDMGEXP/CROPDMGEXP give a multiplier code (per the NWS documentation: K = thousand, M = million, B = billion; numeric digits and a few stray symbols also appear in the raw data and are treated as powers of ten or as unknown/zero respectively, following the commonly used convention for this dataset).

expToMultiplier <- function(exp) {
    exp <- toupper(trimws(exp))
    case_when(
        exp == "K" ~ 1e3,
        exp == "M" ~ 1e6,
        exp == "B" ~ 1e9,
        exp == "H" ~ 1e2,
        exp %in% as.character(0:9) ~ 10^as.numeric(exp),
        TRUE ~ 0  # blank, "-", "+", "?" and other undocumented codes treated as no valid multiplier
    )
}

storm2 <- storm2 %>%
    mutate(
        propMultiplier = expToMultiplier(PROPDMGEXP),
        cropMultiplier = expToMultiplier(CROPDMGEXP),
        propDamageUSD  = PROPDMG * propMultiplier,
        cropDamageUSD  = CROPDMG * cropMultiplier,
        totalDamageUSD = propDamageUSD + cropDamageUSD
    )

Aggregating by event type

We compute, for each event type, the total fatalities, total injuries, and total economic damage (property + crop), then extract the top 10 event types for each impact measure.

healthByEvent <- storm2 %>%
    group_by(EVTYPE) %>%
    summarise(
        totalFatalities = sum(FATALITIES, na.rm = TRUE),
        totalInjuries   = sum(INJURIES, na.rm = TRUE),
        .groups = "drop"
    ) %>%
    mutate(totalHealthImpact = totalFatalities + totalInjuries) %>%
    arrange(desc(totalHealthImpact))

top10Fatalities <- healthByEvent %>% arrange(desc(totalFatalities)) %>% slice(1:10)
top10Injuries   <- healthByEvent %>% arrange(desc(totalInjuries))   %>% slice(1:10)

economicByEvent <- storm2 %>%
    group_by(EVTYPE) %>%
    summarise(
        totalPropDamage = sum(propDamageUSD, na.rm = TRUE),
        totalCropDamage = sum(cropDamageUSD, na.rm = TRUE),
        totalDamage     = sum(totalDamageUSD, na.rm = TRUE),
        .groups = "drop"
    ) %>%
    arrange(desc(totalDamage))

top10Damage <- economicByEvent %>% slice(1:10)

Results

Which event types are most harmful to population health?

The tables below show the ten event types responsible for the most fatalities and the most injuries, respectively, across the full 1950-2011 record.

kable(top10Fatalities %>% select(EVTYPE, totalFatalities),
      col.names = c("Event Type", "Total Fatalities"),
      caption = "Top 10 event types by total fatalities")
Top 10 event types by total fatalities
Event Type Total Fatalities
TORNADO 5633
EXCESSIVE HEAT 1903
FLASH FLOOD 978
HEAT 937
LIGHTNING 816
TSTM WIND 504
FLOOD 470
RIP CURRENT 368
HIGH WIND 248
AVALANCHE 224 
kable(top10Injuries %>% select(EVTYPE, totalInjuries),
      col.names = c("Event Type", "Total Injuries"),
      caption = "Top 10 event types by total injuries")
Top 10 event types by total injuries
Event Type Total Injuries
TORNADO 91346
TSTM WIND 6957
FLOOD 6789
EXCESSIVE HEAT 6525
LIGHTNING 5230
HEAT 2100
ICE STORM 1975
FLASH FLOOD 1777
THUNDERSTORM WIND 1488
HAIL 1361

Figure 1 below presents these same results graphically as a two-panel bar chart, making it easy to compare the leading causes of death and injury side by side.

p1 <- ggplot(top10Fatalities, aes(x = reorder(EVTYPE, totalFatalities), y = totalFatalities)) +
    geom_col(fill = "firebrick") +
    coord_flip() +
    labs(title = "Top 10 Causes of Fatalities", x = "Event Type", y = "Total Fatalities") +
    theme_minimal()

p2 <- ggplot(top10Injuries, aes(x = reorder(EVTYPE, totalInjuries), y = totalInjuries)) +
    geom_col(fill = "darkorange") +
    coord_flip() +
    labs(title = "Top 10 Causes of Injuries", x = "Event Type", y = "Total Injuries") +
    theme_minimal()

grid.arrange(p1, p2, ncol = 2)

Figure 1. Top 10 severe weather event types in the United States (1950-2011) ranked by total fatalities (left panel) and total injuries (right panel). Tornadoes are the leading cause of both fatalities and injuries by a wide margin.

Across the United States, tornadoes are clearly the most harmful event type to population health, causing far more fatalities and injuries than any other category, followed by excessive heat and flash floods/thunderstorm winds for fatalities, and thunderstorm wind/flood-related events for injuries.

Which event types have the greatest economic consequences?

The table below shows the ten event types with the greatest combined property and crop damage.

top10Damage_display <- top10Damage %>%
    mutate(
        `Property Damage ($)` = dollar(totalPropDamage),
        `Crop Damage ($)`     = dollar(totalCropDamage),
        `Total Damage ($)`    = dollar(totalDamage)
    ) %>%
    select(EVTYPE, `Property Damage ($)`, `Crop Damage ($)`, `Total Damage ($)`)

kable(top10Damage_display, col.names = c("Event Type", "Property Damage ($)", "Crop Damage ($)", "Total Damage ($)"),
      caption = "Top 10 event types by total economic damage (property + crop)")
Top 10 event types by total economic damage (property + crop)
Event Type Property Damage ($) Crop Damage ($) Total Damage ($)
FLOOD $144,657,709,800 $5,661,968,450 $150,319,678,250
HURRICANE/TYPHOON $69,305,840,000 $2,607,872,800 $71,913,712,800
TORNADO $56,947,380,614 $414,953,270 $57,362,333,884
STORM SURGE $43,323,536,000 $5,000 $43,323,541,000
HAIL $15,735,267,456 $3,025,954,470 $18,761,221,926
FLASH FLOOD $16,822,723,772 $1,421,317,100 $18,244,040,872
DROUGHT $1,046,106,000 $13,972,566,000 $15,018,672,000
HURRICANE $11,868,319,010 $2,741,910,000 $14,610,229,010
RIVER FLOOD $5,118,945,500 $5,029,459,000 $10,148,404,500
ICE STORM $3,944,927,860 $5,022,113,500 $8,967,041,360

Figure 2 shows the total economic damage (in billions of dollars) for the top 10 event types.

ggplot(top10Damage, aes(x = reorder(EVTYPE, totalDamage), y = totalDamage / 1e9)) +
    geom_col(fill = "steelblue") +
    coord_flip() +
    labs(title = "Top 10 Event Types by Total Economic Damage",
         x = "Event Type", y = "Total Damage (Billions of USD)") +
    theme_minimal()

Figure 2. Top 10 severe weather event types in the United States (1950-2011) ranked by total economic damage (property plus crop damage, in billions of dollars). Floods have caused the greatest cumulative economic damage, with hurricanes/typhoons and tornadoes also among the most costly event types.

Across the United States, floods have caused the greatest total economic damage over the period covered by the database, followed by hurricanes/typhoons and tornadoes. Property damage is the dominant component of total economic loss for most of the top event types, though certain categories (e.g. droughts) contribute disproportionately through crop damage.

Conclusion

Tornadoes represent the single greatest threat to public health among severe weather events in the United States, while floods, hurricanes/typhoons, and tornadoes represent the greatest sources of economic loss. These findings suggest that resource prioritization for severe weather preparedness should weight both the historically high human toll of tornadoes and the substantial economic exposure associated with flooding and tropical storm systems.