Reproducible Research: Peer Assessment 2
SlowLearner
2025-05-27
Severe weather events in the USA and their consequences
Introduction
Storms and other severe weather events can cause both public health
and economic problems for communities and municipalities. Many severe
events can result in fatalities, injuries, and property damage, and
preventing such outcomes to the extent possible is a key concern.
This project involves exploring the U.S. National Oceanic and
Atmospheric Administration’s (NOAA) storm database. This database tracks
characteristics of major storms and weather events in the United States,
including when and where they occur, as well as estimates of any
fatalities, injuries, and property damage.
Synopsis
The objective of our analysis is to use the NOAA storm database to answer the following questions.
Q1. Across the United States, which types of events (as indicated in the EVTYPE variable) are most harmful with respect to population health?
Q2. Across the United States, which types of events have the greatest economic consequences?
Fatalities and injuries variables (FATALITIES,
INJURIES) will be used to identify event types most
harmful to population health (Q1)
Property and crop damage
variables (PROPDMG, CROPDMG) will be used to
identify event types with the greatest economic consequences
(Q2)
We are not expected to qualify our analysis results
either by location or date.
The analysis variables are
measured at 2 different scales:
Fatalities and injuries variables are expressed as a count of individuals affected.
Property and crop damage variables are expressed as dollar amounts.
These variables are normalized up front to ensure they contribute equally to the analysis.
Data Processing and Analysis
Setup
Load required libraries, assign work variables and show session information.
library(ggplot2)
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, unionlibrary(kableExtra)##
## Attaching package: 'kableExtra'## The following object is masked from 'package:dplyr':
##
## group_rowslibrary(gridExtra)##
## Attaching package: 'gridExtra'## The following object is masked from 'package:dplyr':
##
## combinelocal_zip <- "StormData.bz2"
local_csv <- "StormData.csv"
source_url <- "https://d396qusza40orc.cloudfront.net/repdata%2Fdata%2FStormData.csv.bz2"
data_dir <- "./Data"
sessionInfo()## R version 4.3.3 (2024-02-29 ucrt)
## Platform: x86_64-w64-mingw32/x64 (64-bit)
## Running under: Windows 10 x64 (build 19045)
##
## Matrix products: default
##
##
## locale:
## [1] LC_COLLATE=English_South Africa.utf8 LC_CTYPE=English_South Africa.utf8
## [3] LC_MONETARY=English_South Africa.utf8 LC_NUMERIC=C
## [5] LC_TIME=English_South Africa.utf8
##
## time zone: Africa/Johannesburg
## tzcode source: internal
##
## attached base packages:
## [1] stats graphics grDevices utils datasets methods base
##
## other attached packages:
## [1] gridExtra_2.3 kableExtra_1.4.0 dplyr_1.1.4 ggplot2_3.5.1
##
## loaded via a namespace (and not attached):
## [1] gtable_0.3.5 jsonlite_1.8.8 compiler_4.3.3 tidyselect_1.2.1
## [5] xml2_1.3.6 stringr_1.5.1 jquerylib_0.1.4 systemfonts_1.2.2
## [9] scales_1.3.0 yaml_2.3.8 fastmap_1.2.0 R6_2.5.1
## [13] generics_0.1.3 knitr_1.47 tibble_3.2.1 munsell_0.5.1
## [17] svglite_2.1.3 bslib_0.7.0 pillar_1.9.0 rlang_1.1.3
## [21] utf8_1.2.4 cachem_1.1.0 stringi_1.8.3 xfun_0.44
## [25] sass_0.4.9 viridisLite_0.4.2 cli_3.6.2 withr_3.0.0
## [29] magrittr_2.0.3 digest_0.6.35 grid_4.3.3 rstudioapi_0.16.0
## [33] lifecycle_1.0.4 vctrs_0.6.5 evaluate_0.24.0 glue_1.7.0
## [37] fansi_1.0.6 colorspace_2.1-0 rmarkdown_2.27 tools_4.3.3
## [41] pkgconfig_2.0.3 htmltools_0.5.8.1Data preparation
Retrieve source data
The source data for analysis is a compressed comma separated values (.csv) located at https://d396qusza40orc.cloudfront.net/repdata%2Fdata%2FStormData.csv.bz2.
Load data into R environment
- The file is downloaded to a local file folder.
local_zip <- paste(data_dir, local_zip, sep='/')
if (!dir.exists(data_dir)) dir.create(data_dir)
if(!file.exists(local_zip)) {
download.file(source_url, destfile=local_zip , mode='wb')
}- Create the analysis input dataset
local_csvusing theread.tablecommand (no need need to unzip the compressed file separately, this command handles automatically) - Provide some basic stats about the file created.
storms_data <- read.table(local_zip, sep=',', header=TRUE)Loaded analysis data overview
typ <- class(storms_data)
nro <- nrow(storms_data)
nva <- ncol(storms_data)
icv <- complete.cases(storms_data)
nic <- nro - length(icv[icv == FALSE])storms_datais of type data.frame.- It contains 902297 observations.
- Each observation has 37 variables.
- There are 0 incomplete cases in the data
- Compact structure display
str(storms_data)## 'data.frame': 902297 obs. of 37 variables:
## $ STATE__ : num 1 1 1 1 1 1 1 1 1 1 ...
## $ 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" ...
## $ BGN_TIME : chr "0130" "0145" "1600" "0900" ...
## $ TIME_ZONE : chr "CST" "CST" "CST" "CST" ...
## $ COUNTY : num 97 3 57 89 43 77 9 123 125 57 ...
## $ COUNTYNAME: chr "MOBILE" "BALDWIN" "FAYETTE" "MADISON" ...
## $ STATE : chr "AL" "AL" "AL" "AL" ...
## $ EVTYPE : chr "TORNADO" "TORNADO" "TORNADO" "TORNADO" ...
## $ BGN_RANGE : num 0 0 0 0 0 0 0 0 0 0 ...
## $ BGN_AZI : chr "" "" "" "" ...
## $ BGN_LOCATI: chr "" "" "" "" ...
## $ END_DATE : chr "" "" "" "" ...
## $ END_TIME : chr "" "" "" "" ...
## $ COUNTY_END: num 0 0 0 0 0 0 0 0 0 0 ...
## $ COUNTYENDN: logi NA NA NA NA NA NA ...
## $ END_RANGE : num 0 0 0 0 0 0 0 0 0 0 ...
## $ END_AZI : chr "" "" "" "" ...
## $ END_LOCATI: chr "" "" "" "" ...
## $ LENGTH : num 14 2 0.1 0 0 1.5 1.5 0 3.3 2.3 ...
## $ WIDTH : num 100 150 123 100 150 177 33 33 100 100 ...
## $ F : int 3 2 2 2 2 2 2 1 3 3 ...
## $ MAG : num 0 0 0 0 0 0 0 0 0 0 ...
## $ 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 "" "" "" "" ...
## $ WFO : chr "" "" "" "" ...
## $ STATEOFFIC: chr "" "" "" "" ...
## $ ZONENAMES : chr "" "" "" "" ...
## $ LATITUDE : num 3040 3042 3340 3458 3412 ...
## $ LONGITUDE : num 8812 8755 8742 8626 8642 ...
## $ LATITUDE_E: num 3051 0 0 0 0 ...
## $ LONGITUDE_: num 8806 0 0 0 0 ...
## $ REMARKS : chr "" "" "" "" ...
## $ REFNUM : num 1 2 3 4 5 6 7 8 9 10 ...Discard noise
Remove attributes not required for the analysis and free unused memory.
storms_abbr <- select(storms_data,c('EVTYPE',
'FATALITIES',
'INJURIES',
'PROPDMG',
'CROPDMG'))
str(storms_abbr)## 'data.frame': 902297 obs. of 5 variables:
## $ 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 ...
## $ CROPDMG : num 0 0 0 0 0 0 0 0 0 0 ...gc()## used (Mb) gc trigger (Mb) max used (Mb)
## Ncells 1439012 76.9 2503190 133.7 1662101 88.8
## Vcells 60519346 461.8 177304376 1352.8 221630414 1691.0Data Analysis
The prepared file storms_abbr is processed by:
- normalizing the data
- grouping by event type
EVTYPE - summarizing to provide totals of each variable under scrutiny
Results are stored in results_data.
normalize <- function(x) {
(x - min(x)) / (max(x) - min(x))
}
norm_data <- as.data.frame(lapply(storms_abbr[2:5], normalize))
norm_data$EVTYPE <- storms_abbr$EVTYPE
summary_data <- summarize(group_by(norm_data, EVTYPE),
totFATALITIES = sum(FATALITIES),
totINJURIES = sum(INJURIES),
totPROPDMG = sum(PROPDMG),
totCROPDMG = sum(CROPDMG))
summary(summary_data)## EVTYPE totFATALITIES totINJURIES totPROPDMG
## Length:985 Min. :0.00000 Min. : 0.00000 Min. : 0.000
## Class :character 1st Qu.:0.00000 1st Qu.: 0.00000 1st Qu.: 0.000
## Mode :character Median :0.00000 Median : 0.00000 Median : 0.000
## Mean :0.02637 Mean : 0.08392 Mean : 2.210
## 3rd Qu.:0.00000 3rd Qu.: 0.00000 3rd Qu.: 0.007
## Max. :9.66209 Max. :53.73294 Max. :642.452
## totCROPDMG
## Min. : 0.000
## 1st Qu.: 0.000
## Median : 0.000
## Mean : 1.413
## 3rd Qu.: 0.000
## Max. :585.451results_data <- summary_data %>%
filter(totFATALITIES == max(totFATALITIES) |
totINJURIES == max(totINJURIES) |
totPROPDMG == max(totPROPDMG) |
totCROPDMG == max(totCROPDMG))Plots
Terms used in the plots
- Consequence - umbrella term for the variables under
scrutiny:
FATALITIES,INJURIES,PROPDMG,CROPDMG - Severity - the normalized totals of each consequence
- Peril - synonym for storm / weather event type
summary_data is transformed into a list of the top 5
event types for each variable under scrutiny
(top5_data).
The plots below show storm and other weather events by Peril and categorized by Consequence for the top 5 event types in terms of severity.
top5_data <- bind_rows(
mutate(select(arrange(summary_data, desc(totFATALITIES)),
count = totFATALITIES,
event_type = EVTYPE),
category = "FATALITIES")[1:5,],
mutate(select(arrange(summary_data, desc(totINJURIES)),
count = totINJURIES,
event_type = EVTYPE),
category = "INJURIES")[1:5,],
mutate(select(arrange(summary_data, desc(totPROPDMG)),
count = totPROPDMG,
event_type = EVTYPE),
category = "PROPDMG")[1:5,],
mutate(select(arrange(summary_data, desc(totCROPDMG)),
count = totCROPDMG,
event_type = EVTYPE),
category = "CROPDMG")[1:5,])
p1 <- ggplot(data=top5_data[top5_data$category %in% c("FATALITIES", "INJURIES"),],
aes(fill=category, y=count, x=event_type)) +
geom_bar(stat="identity", width=.9) +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1),
plot.margin = margin(0,0.0,0.2,0.7, "cm")) +
labs(x = "Peril",
y = "Severity",
title = "Health outcomes") +
scale_fill_brewer(palette=4, direction=-1)
p2 <- ggplot(data=top5_data[top5_data$category %in% c("PROPDMG", "CROPDMG"),],
aes(fill=category, y=count, x=event_type)) +
geom_bar(stat="identity", width=.9) +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1),
plot.margin = margin(0,0.0,0.2,0.7, "cm")) +
labs(x = "Peril",
y = "Severity",
title = "Economic impact") +
scale_fill_brewer(palette="Set3", direction=-1)
grid.arrange(grobs = list(p1, p2), ncol = 2, top="Figure 1: Severities by Peril")
The data of top5_data (see table below) shows most
clearly that i) EXCESSIVE HEAT is amongst the top 5 health
consequences, and ii) FLASH FLOODS are the 2nd
most severe economic consequence for both property and crop damage.
Further analysis could be done to see how these have changed since measurements began.
select(top5_data, category, event_type, count) %>%
group_by(category) %>%
kbl(caption = "Table 1: Top 5 severe weather event types") %>%
kable_material(c("basic")) %>%
kable_styling(bootstrap_options = "striped",
font_size = 12) %>%
collapse_rows(valign = "top")| category | event_type | count |
|---|---|---|
| FATALITIES | TORNADO | 9.662093 |
| EXCESSIVE HEAT | 3.264151 | |
| FLASH FLOOD | 1.677530 | |
| HEAT | 1.607204 | |
| LIGHTNING | 1.399657 | |
| INJURIES | TORNADO | 53.732941 |
| TSTM WIND | 4.092353 | |
| FLOOD | 3.993529 | |
| EXCESSIVE HEAT | 3.838235 | |
| LIGHTNING | 3.076471 | |
| PROPDMG | TORNADO | 642.451632 |
| FLASH FLOOD | 284.024918 | |
| TSTM WIND | 267.193122 | |
| FLOOD | 179.987696 | |
| THUNDERSTORM WIND | 175.368834 | |
| CROPDMG | HAIL | 585.450788 |
| FLASH FLOOD | 181.010566 | |
| FLOOD | 169.735232 | |
| TSTM WIND | 110.305657 | |
| TORNADO | 101.028808 |
Results
The table below shows the event types that have the severest impact in terms of:
Health (Q1)
- fatalities
- injuries
Economic consequences (Q2)
- property damage
- crop damage
colnames(results_data) <- c("Event Type", "Total Fatalities", "Total Injuries", "Total Property Damage", "Total Crop Damage") results_data %>% kbl(caption = "Table 2: Weather events with severest impact") %>% add_header_above(c(" ", "Health" = 2, "Economic" = 2)) %>% kable_styling(bootstrap_options = "striped", font_size = 12) %>% kable_material(c("basic"))Table 2: Weather events with severest impact
Event Type
Total Fatalities
Total Injuries
Total Property Damage
Total Crop Damage
HAIL
0.025729
0.8005882
137.7387
585.4508
TORNADO
9.662093
53.7329412
642.4516
101.0288
Conclusion
Tornadoes are the most destructive weather events regarding fatalities, injuries and property damage.
Hail storms do the most damage to crops. Although tornadoes also wreak havoc to crops, the economic impact of hail storms is close to six times greater!
—– End —–