Severe Weather Impact on US Health and Economics

Synopsis

This project explores the U.S. National Oceanic and Atmospheric Administration’s (NOAA) storm database, which tracks weather events and major storms, to evaluate the most severe types of weather events in the USA. Specifically, I want to address which catastrophic events caused the greatest damage to the human population (fatalities / injuries) and that cause economic loss (property / crop damage) between the years 1950 - 2011.

This analysis have two major goal -

To identify which kind of weather events are the most harmful for the health of the population, and
To identify which have the highest negative consequences on the economy.

Data Processing

The data can be downloaded from here and read by the read.csv() function.

stormData <- read.csv("repdata_data_StormData.csv.bz2", header = T)

Up next, the libraries needed for analysis are loaded.

library(lubridate)
library(plyr)
library(dplyr)
library(ggplot2)
library(gridExtra)
library(magrittr)

Then, column names are checked by names() function.

names(stormData)

##  [1] "STATE__"    "BGN_DATE"   "BGN_TIME"   "TIME_ZONE"  "COUNTY"    
##  [6] "COUNTYNAME" "STATE"      "EVTYPE"     "BGN_RANGE"  "BGN_AZI"   
## [11] "BGN_LOCATI" "END_DATE"   "END_TIME"   "COUNTY_END" "COUNTYENDN"
## [16] "END_RANGE"  "END_AZI"    "END_LOCATI" "LENGTH"     "WIDTH"     
## [21] "F"          "MAG"        "FATALITIES" "INJURIES"   "PROPDMG"   
## [26] "PROPDMGEXP" "CROPDMG"    "CROPDMGEXP" "WFO"        "STATEOFFIC"
## [31] "ZONENAMES"  "LATITUDE"   "LONGITUDE"  "LATITUDE_E" "LONGITUDE_"
## [36] "REMARKS"    "REFNUM"

Three things to be modified for analysis -

BGN_DATE column is changed to date type with mdy_hms() function.
Based on the information present in the NOAA, the data started to be recorder in January 1950, but at that time only tornado event were registered. With the passage of time more kind of weather events were added, until, in January 1996, all event type started to get record. Due to the fact that one of my main goal is to compare the effect of different weather events, I will firstly subset our set of data for all the event recorded from January 1996 onward with subset() function.
From the column names, only EVTYPE, FATALITIES, INJURIES, PROPDMG, PROPDMGEXP, CROPDMG, CROPDMGEXP are of necessary. These will be in the select parameter of the subset() function.

stormData$BGN_DATE <- mdy_hms(stormData$BGN_DATE)
stormData_sub <- subset(stormData, BGN_DATE > "1995-12-31", 
                        select = c(EVTYPE, FATALITIES, INJURIES, PROPDMG, PROPDMGEXP, CROPDMG, CROPDMGEXP))

Results

Question 1:

Across the United States, which types of events (as indicated in the EVTYPE variable) are most harmful with respect to population health?

The FATALITIES and INJURIES variables are to be aggregated seperately with the aggregate() function, and sum will be it’s FUN parameter.

totalFatality <- aggregate(FATALITIES ~ EVTYPE, data = stormData_sub, FUN = sum)
totalInjury <- aggregate(INJURIES ~ EVTYPE, data = stormData_sub, FUN = sum)

The aggregated data is arranged in a descending order base on the EVTYPE variable with the arrange() function, and top 10 are selected for analysis.

stormDataFatality <- arrange(totalFatality, desc(FATALITIES), EVTYPE)[1:10, ]
show(stormDataFatality)

##            EVTYPE FATALITIES
## 1  EXCESSIVE HEAT       1797
## 2         TORNADO       1511
## 3     FLASH FLOOD        887
## 4       LIGHTNING        651
## 5           FLOOD        414
## 6     RIP CURRENT        340
## 7       TSTM WIND        241
## 8            HEAT        237
## 9       HIGH WIND        235
## 10      AVALANCHE        223

stormDataInjury <- arrange(totalInjury, desc(INJURIES), EVTYPE)[1:10, ]
show(stormDataInjury)

##               EVTYPE INJURIES
## 1            TORNADO    20667
## 2              FLOOD     6758
## 3     EXCESSIVE HEAT     6391
## 4          LIGHTNING     4141
## 5          TSTM WIND     3629
## 6        FLASH FLOOD     1674
## 7  THUNDERSTORM WIND     1400
## 8       WINTER STORM     1292
## 9  HURRICANE/TYPHOON     1275
## 10              HEAT     1222

Now, these data are plotted side by side with the ggplot() function, and the grid.arrange() function.

plotFatality <- ggplot(data = stormDataFatality, aes(x = reorder(EVTYPE, -FATALITIES), y = FATALITIES)) +
    geom_bar(stat = "identity", fill = "red3") + 
    theme(axis.text.x = element_text(angle = 40, hjust = 1)) +
    xlab("") +  ylab("Total number of fatalities")

plotInjury <- ggplot(data = stormDataInjury, aes(x = reorder(EVTYPE, -INJURIES), y = INJURIES)) +
    geom_bar(stat = "identity", fill = "red3") + 
    theme(axis.text.x = element_text(angle = 40, hjust = 1)) +
    xlab("") +  ylab("Total number of injuries")

grid.arrange(plotFatality, plotInjury, ncol = 2, 
             top = "Fatalities & Injuries from top 10 Weather Events - Public Health Impact")

The histogram shows that Tornado’s are the most harmful weather events for people’s fatalities / injuries.

Question 2:

Across the United States, which types of events have the greatest economic consequences?

First, unique values of the PROPDMGEXP and CROPDMGEXP are checked.

unique(stormData_sub$PROPDMGEXP)

## [1] ""  "M" "K" "B" "0"

unique(stormData_sub$CROPDMGEXP)

## [1] ""  "M" "K" "B"

The components of PROPDMGEXP and CROPDMGEXP variables need to be converted into a common exponential number by mapvalues() function, and will be multiplied with the values of PROPDMG and CROPDMG to get total damage.

stormData_sub$totalPROPDMG <- (mapvalues(stormData_sub$PROPDMGEXP, from = c("", "K", "M", "B"), 
                                         to = c(1, 1e3, 1e6, 1e9)) %>% as.numeric) * stormData_sub$PROPDMG
stormData_sub$totalCROPDMG <- (mapvalues(stormData_sub$CROPDMGEXP, from = c("", "K", "M", "B"), 
                                         to = c(1, 1e3, 1e6, 1e9)) %>% as.numeric) * stormData_sub$CROPDMG

Then, these are to be aggregated separately with the aggregate() function, and sum will be it’s FUN parameter.

totalPropDmg <- aggregate(totalPROPDMG ~ EVTYPE, data = stormData_sub, FUN = sum)
totalCropDmg <- aggregate(totalCROPDMG ~ EVTYPE, data = stormData_sub, FUN = sum)