Analysis of Severe Weather Events in the United States (1950

Synopsis

Severe weather events can cause significant loss of life and economic damage. Understanding which types of events have the greatest impact helps authorities and policymakers prioritize resources for prevention and response. This analysis uses the NOAA Storm Database (1950–2011) to identify the weather events most harmful to population health and those with the highest economic consequences across the United States. Tornadoes were found to be the most harmful to human health, while floods and hurricanes resulted in the greatest economic losses. All steps, including data loading, cleaning, and analysis, are fully reproducible from the raw data file.

Data Processing

The raw NOAA Storm Database is provided as a compressed CSV (.csv.bz2). All preprocessing steps are performed within this document, starting from downloading the raw file. Damage estimates use the exponents (PROPDMGEXP, CROPDMGEXP) to convert values into USD. Event types (EVTYPE) are standardized to uppercase for consistency.

library(dplyr)

## 
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

library(readr)
library(stringr)
library(ggplot2)
library(scales)

## 
## Attaching package: 'scales'

## The following object is masked from 'package:readr':
## 
##     col_factor

library(knitr)

# Download raw data
url <- "https://d396qusza40orc.cloudfront.net/repdata%2Fdata%2FStormData.csv.bz2"
destfile <- "StormData.csv.bz2"
if(!file.exists(destfile)) download.file(url, destfile, mode = "wb")

storm <- read_csv(destfile)

## Rows: 902297 Columns: 37

## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (18): BGN_DATE, BGN_TIME, TIME_ZONE, COUNTYNAME, STATE, EVTYPE, BGN_AZI,...
## dbl (18): STATE__, COUNTY, BGN_RANGE, COUNTY_END, END_RANGE, LENGTH, WIDTH, ...
## lgl  (1): COUNTYENDN
## 
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

# Function to convert exponent to multiplier
exp_to_mult <- function(exp) {
  if(is.na(exp) || exp == "") return(1)
  e <- toupper(str_trim(as.character(exp)))
  if(e == "H") return(100)
  if(e == "K") return(1e3)
  if(e == "M") return(1e6)
  if(e == "B") return(1e9)
  return(1)
}

df <- storm %>%
  select(EVTYPE, FATALITIES, INJURIES, PROPDMG, PROPDMGEXP, CROPDMG, CROPDMGEXP) %>%
  mutate(
    EVTYPE = str_to_upper(str_trim(EVTYPE)),
    PROP_MULT = sapply(PROPDMGEXP, exp_to_mult),
    CROP_MULT = sapply(CROPDMGEXP, exp_to_mult),
    PROP_DMG_USD = coalesce(PROPDMG,0) * PROP_MULT,
    CROP_DMG_USD = coalesce(CROPDMG,0) * CROP_MULT,
    TOTAL_ECONOMIC = PROP_DMG_USD + CROP_DMG_USD,
    TOTAL_HEALTH = coalesce(FATALITIES,0) + coalesce(INJURIES,0)
  )

Results

1. Events Most Harmful to Population Health

by_event <- df %>%
  group_by(EVTYPE) %>%
  summarise(
    FATALITIES = sum(FATALITIES, na.rm = TRUE),
    INJURIES = sum(INJURIES, na.rm = TRUE),
    HEALTH_IMPACT = sum(TOTAL_HEALTH, na.rm = TRUE),
    PROP_DAMAGE = sum(PROP_DMG_USD, na.rm = TRUE),
    CROP_DAMAGE = sum(CROP_DMG_USD, na.rm = TRUE),
    ECON_IMPACT = sum(TOTAL_ECONOMIC, na.rm = TRUE)
  )

top_health <- by_event %>% arrange(desc(HEALTH_IMPACT)) %>% slice(1:10)
kable(top_health[, c("EVTYPE","FATALITIES","INJURIES","HEALTH_IMPACT")], caption = "Top 10 Event Types by Health Impact")

Top 10 Event Types by Health Impact
EVTYPE	FATALITIES	INJURIES	HEALTH_IMPACT
TORNADO	5633	91346	96979
EXCESSIVE HEAT	1903	6525	8428
TSTM WIND	504	6957	7461
FLOOD	470	6789	7259
LIGHTNING	816	5230	6046
HEAT	937	2100	3037
FLASH FLOOD	978	1777	2755
ICE STORM	89	1975	2064
THUNDERSTORM WIND	133	1488	1621
WINTER STORM	206	1321	1527

ggplot(top_health, aes(x=reorder(EVTYPE, HEALTH_IMPACT), y=HEALTH_IMPACT)) +
  geom_col(fill="steelblue") +
  coord_flip() +
  labs(x="Event Type", y="Total Fatalities + Injuries",
       title="Most Harmful Weather Events to Population Health") +
  theme_minimal()

Figure 1: Top 10 severe weather events causing the highest combined fatalities and injuries across the United States (1950–2011).

Interpretation: Tornadoes cause the most injuries and fatalities combined, followed by excessive heat, floods, and lightning.

2. Events with Greatest Economic Consequences

top_econ <- by_event %>% arrange(desc(ECON_IMPACT)) %>% slice(1:10)
kable(top_econ[, c("EVTYPE","PROP_DAMAGE","CROP_DAMAGE","ECON_IMPACT")], caption = "Top 10 Event Types by Economic Damage")

Top 10 Event Types by Economic Damage
EVTYPE	PROP_DAMAGE	CROP_DAMAGE	ECON_IMPACT
FLOOD	144657709807	5661968450	150319678257
HURRICANE/TYPHOON	69305840000	2607872800	71913712800
TORNADO	56937160779	414953270	57352114049
STORM SURGE	43323536000	5000	43323541000
HAIL	15732267543	3025954473	18758222016
FLASH FLOOD	16140862067	1421317100	17562179167
DROUGHT	1046106000	13972566000	15018672000
HURRICANE	11868319010	2741910000	14610229010
RIVER FLOOD	5118945500	5029459000	10148404500
ICE STORM	3944927860	5022113500	8967041360

ggplot(top_econ, aes(x=reorder(EVTYPE, ECON_IMPACT), y=ECON_IMPACT/1e9)) +
  geom_col(fill="darkred") +
  coord_flip() +
  labs(x="Event Type", y="Total Economic Damage (Billions USD)",
       title="Weather Events with Greatest Economic Consequences") +
  theme_minimal()

Figure 2: Top 10 severe weather events causing the largest combined property and crop damages in the United States (1950–2011).

Interpretation: Floods and hurricanes/typhoons result in the largest total economic losses.

Conclusion

Most harmful to health: Tornadoes
Greatest economic impact: Floods and hurricanes
All transformations and calculations are reproducible from the raw data file.

sessionInfo()

## R version 4.4.2 (2024-10-31 ucrt)
## Platform: x86_64-w64-mingw32/x64
## Running under: Windows 11 x64 (build 22631)
## 
## Matrix products: default
## 
## 
## locale:
## [1] LC_COLLATE=Dutch_Netherlands.utf8  LC_CTYPE=Dutch_Netherlands.utf8   
## [3] LC_MONETARY=Dutch_Netherlands.utf8 LC_NUMERIC=C                      
## [5] LC_TIME=Dutch_Netherlands.utf8    
## 
## time zone: Europe/Amsterdam
## tzcode source: internal
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
## [1] knitr_1.49    scales_1.3.0  ggplot2_3.5.1 stringr_1.5.1 readr_2.1.5  
## [6] dplyr_1.1.4  
## 
## loaded via a namespace (and not attached):
##  [1] bit_4.5.0.1       gtable_0.3.6      jsonlite_1.8.9    crayon_1.5.3     
##  [5] compiler_4.4.2    tidyselect_1.2.1  parallel_4.4.2    jquerylib_0.1.4  
##  [9] yaml_2.3.10       fastmap_1.2.0     R6_2.6.1          labeling_0.4.3   
## [13] generics_0.1.3    tibble_3.2.1      munsell_0.5.1     bslib_0.9.0      
## [17] pillar_1.10.1     tzdb_0.4.0        rlang_1.1.4       cachem_1.1.0     
## [21] stringi_1.8.4     xfun_0.50         sass_0.4.9        bit64_4.5.2      
## [25] cli_3.6.3         withr_3.0.2       magrittr_2.0.3    digest_0.6.37    
## [29] grid_4.4.2        vroom_1.6.5       rstudioapi_0.17.1 hms_1.1.3        
## [33] lifecycle_1.0.4   vctrs_0.6.5       evaluate_1.0.3    glue_1.8.0       
## [37] farver_2.1.2      codetools_0.2-20  colorspace_2.1-1  rmarkdown_2.29   
## [41] tools_4.4.2       pkgconfig_2.0.3   htmltools_0.5.8.1

Analysis of Severe Weather Events in the United States (1950–2011)