Analyzing Natural Events causing Public Health and Economic Problems in U.S.
MOHAMMAD SHADAN
September 5, 2016
SYNOPSIS
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.
Data analysis address the below concerns:
Across the United States, which types of events (as indicated in the EVTYPE variable) are most harmful with respect to population health?
Across the United States, which types of events have the greatest economic consequences
ASSUMPTIONS ON WHICH ANALYSIS IS BASED
Population Health is affected by “FATALITIES” and “INJURIES”
Economic Consequences are because of Property Damage (“PROPDMG”) and Crop Damage (“CROPDMG”)
DATA PROCESSING
Storm Data can be download from the URL
Set the Working Directory (e.g. setwd(“~/R/RR”)
Move the StormData.csv file in to Working Directory before the script is run
#Install the Required Packages
#install.packages("dplyr"); install.packages("ggplot2"); install.packages("gridExtra")
#Loading Requierd Libraries
library(dplyr); library(ggplot2); library(gridExtra);##
## 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##
## Attaching package: 'gridExtra'## The following object is masked from 'package:dplyr':
##
## combine#Extracting data file StormData.csv
s <- read.csv("StormData.csv",sep = ",", header = TRUE)#Number of Columns and Rows in the StormData.csv database
dim(s)## [1] 902297 37#Name of the Cloumns in the StormData.csv datase
names(s)## [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"For this analysis we will be consdering Seven Columns as below. The respective column numbers are mentioned in brackets
ETYTPE(8), FATALITIES(23), INJURIES(24), PROPDMG(25), PROPDMGEXP(26), CROPDMG(27), CROPDMGEXP(28)
#Fetching only the required Columns
s<-s[,c(8,23:28)]
head(s)## EVTYPE FATALITIES INJURIES PROPDMG PROPDMGEXP CROPDMG CROPDMGEXP
## 1 TORNADO 0 15 25.0 K 0
## 2 TORNADO 0 0 2.5 K 0
## 3 TORNADO 0 2 25.0 K 0
## 4 TORNADO 0 2 2.5 K 0
## 5 TORNADO 0 2 2.5 K 0
## 6 TORNADO 0 6 2.5 K 0- Across the United States, which types of events (as indicated in the EVTYPE variable) are most harmful with respect to population health?
Data Processing Related to Population Health
Considering that Population Health is affected by “FATALITIES” and “INJURIES”
##Data Processing for FATALITIES (one of reasons affecting Poplulation Health)
agg_fatalities <- aggregate(FATALITIES~EVTYPE, s, FUN=sum, na.rm=TRUE)
agg_fatalities <- arrange(agg_fatalities, desc(FATALITIES))
agg_fatalities5 <- agg_fatalities[1:5,]
head(agg_fatalities5)## EVTYPE FATALITIES
## 1 TORNADO 5633
## 2 EXCESSIVE HEAT 1903
## 3 FLASH FLOOD 978
## 4 HEAT 937
## 5 LIGHTNING 816#Script to plot Total Fatalities with respect to Event Type
ft <- ggplot(agg_fatalities5, aes(x=reorder(EVTYPE, -FATALITIES), y=FATALITIES,FILL=EVTYPE))
ft <- ft + geom_bar(aes(fill=EVTYPE),stat="identity")
ft <- ft + geom_text(aes(label=round(FATALITIES,0), hjust=0.5, vjust=-0.7), size=3.5)
ft <- ft + labs(x = "\n EVENT TYPE \n")
ft <- ft + labs(y = "\n FATALITIES \n")
ft <- ft + labs(title = "\n EVENTS CAUSING FATALITIES \n (TOP 5)\n")
ft <- ft + coord_cartesian(ylim=c(0,6000))
ft <- ft + theme(axis.text.x = element_text(angle=45, hjust=1, size=10, face = "bold"))
ft <- ft + theme(legend.position="none")
ft <- ft + theme(plot.title = element_text(size=11, face = "bold"))##Data Processing for INJURIES (one of reasons affecting Poplulation Health)
agg_injuries <- aggregate(INJURIES~EVTYPE, s, FUN=sum, na.rm=TRUE)
agg_injuries <- arrange(agg_injuries, desc(INJURIES))
agg_injuries5 <- agg_injuries[1:5,]
head(agg_injuries5)## EVTYPE INJURIES
## 1 TORNADO 91346
## 2 TSTM WIND 6957
## 3 FLOOD 6789
## 4 EXCESSIVE HEAT 6525
## 5 LIGHTNING 5230###Script to plot Total Injuries with respect to Event Type
ij <- ggplot(agg_injuries5, aes(x=reorder(EVTYPE, -INJURIES), y=INJURIES,FILL=EVTYPE))
ij <- ij + geom_bar(aes(fill=EVTYPE),stat="identity")
ij <- ij + geom_text(aes(label=round(INJURIES,0), hjust=0.5, vjust=-0.7),size=3.5)
ij <- ij + labs(x="\n EVENT TYPE \n")
ij <- ij + labs(y="\n INJURIES \n")
ij <- ij + labs(title = "\n EVENTS CAUSING INJURIES \n (TOP 5)\n")
ij <- ij + coord_cartesian(ylim=c(0,100000))
ij <- ij + theme(axis.text.x = element_text(angle=45, hjust=1, size=10, face = "bold"))
ij <- ij + theme(legend.position="none")
ij <- ij + theme(plot.title = element_text(size=11, face = "bold"))#Panel Plot for Events affecting Population Health
grid.arrange(ft, ij, ncol=2)
###Data Processing to merge the data sets of Total Fatalities and Total Injuries
merge_health <- merge(agg_fatalities, agg_injuries, all=TRUE)
merge_health[is.na(merge_health)] <- 0
merge_health <- mutate(merge_health, TOTAL = FATALITIES + INJURIES)
merge_health <- arrange(merge_health, desc(TOTAL))#Top Five Events affecting Public Health (Fatalities and Injuries)
head(merge_health,5)## EVTYPE FATALITIES INJURIES TOTAL
## 1 TORNADO 5633 91346 96979
## 2 EXCESSIVE HEAT 1903 6525 8428
## 3 TSTM WIND 504 6957 7461
## 4 FLOOD 470 6789 7259
## 5 LIGHTNING 816 5230 6046#Fetching Top 5 Events
merge_health <- merge_health[1:5,]
###Script to plot Total Fatalities and Injuries with respect to Event Type
ph <- ggplot(merge_health, aes(x=reorder(EVTYPE, -TOTAL), y=TOTAL))
ph <- ph + geom_bar(aes(fill=EVTYPE),stat="identity")
ph <- ph + geom_text(aes(label=round(TOTAL,0), hjust=0.5, vjust=-0.7),size=3.5)
ph <- ph + labs(x="\n EVENT TYPE \n")
ph <- ph + labs(y="\n TOTAL AFFECTED\n")
ph <- ph + labs(title = "\n Top 5 Events Affecting Public Health \n (Considering FATALITIES and INJURIES together)\n")
ph <- ph + coord_cartesian(ylim=c(0,110000))
ph <- ph + theme(axis.text.x = element_text(angle=45, hjust=1, size=12, face = "bold"))
ph <- ph + theme(legend.position="none")
ph <- ph + theme(plot.title = element_text(size=14, face = "bold"))
#Plot for Top 5 Events Affecting Public Health
ph
- Across the United States, which types of events have the greatest economic consequences
Data Processing related to Economic Consequences
Considering economic Consequences are because of Property Damage(“PROPDMG”) and Crop Damage(“CROPDMG”)
Page 12 of Storm Data Documentation states " Alphabetical characters used to signify magnitude include “K” for thousands, “M” for millions, and “B” for billions (in the columns PROPDMGEXP, CROPDMGEXP“)
Assuming Lower Case Letter (e.g. m) means the same as Upper Case Letters (e.g. M)
Update “h” and “H” to 100 (10^2) (Though not mentioned in the Document but assuming “h”, “H” means Hundred)
Update “k” and “K” to 1000 (10^3)
Update “m” and “M” to 1,000,000 (10^6)
Update “b” and “B” to 1,000,000,000 (10^9)
In Storm Data Documentation there are no details to show what “-”, “?”, “+” stands for so I chose to ignore them in the calculations
#Data Processing for Property Damage (PROPDMG)
#List of characters in the columns PROPDMGEXP
table(s$PROPDMGEXP)##
## - ? + 0 1 2 3 4 5
## 465934 1 8 5 216 25 13 4 4 28
## 6 7 8 B h H K m M
## 4 5 1 40 1 6 424665 7 11330p <- filter(s, PROPDMGEXP %in% c("h", "H", "k", "K","m","M","b", "B"))
table(p$PROPDMGEXP)##
## - ? + 0 1 2 3 4 5
## 0 0 0 0 0 0 0 0 0 0
## 6 7 8 B h H K m M
## 0 0 0 40 1 6 424665 7 11330#Replacing "h", "H", "k", "K","m","M","b", "B" by their numerical counterparts
p$PROPDMG <- ifelse((p$PROPDMGEXP =="h"| p$PROPDMGEXP =="H"),p$PROPDMG*10^2,p$PROPDMG)
p$PROPDMG <- ifelse((p$PROPDMGEXP =="k"| p$PROPDMGEXP =="K"),p$PROPDMG*10^3,p$PROPDMG)
p$PROPDMG <- ifelse((p$PROPDMGEXP =="m"| p$PROPDMGEXP =="M"),p$PROPDMG*10^6,p$PROPDMG)
p$PROPDMG <- ifelse((p$PROPDMGEXP =="b"| p$PROPDMGEXP =="B"),p$PROPDMG*10^9,p$PROPDMG)
#Aggrgating Total Property Damage by Event Type
agg_propdmg <- aggregate(PROPDMG~EVTYPE, p, FUN=sum, na.rm=TRUE)
#Sorting the data so that Event causing Maximum Property Damage is on Top
agg_propdmg <- arrange(agg_propdmg, desc(PROPDMG))
#Selecting the top 5 Events causing maximum Property Damage
agg_propdmg5 <- agg_propdmg[1:5,]#Top Five Events Causing Property Damage
head(agg_propdmg5)## EVTYPE PROPDMG
## 1 FLOOD 144657709800
## 2 HURRICANE/TYPHOON 69305840000
## 3 TORNADO 56937160480
## 4 STORM SURGE 43323536000
## 5 FLASH FLOOD 16140811510#Data Processing for Crop Damage (CROPDMG)
#List of characters in the columns CROPDMGEXP
table(s$CROPDMGEXP)##
## ? 0 2 B k K m M
## 618413 7 19 1 9 21 281832 1 1994c <- filter(s, CROPDMGEXP %in% c("h", "H", "k", "K","m","M","b", "B"))
#Replacing "h", "H", "k", "K","m","M","b", "B" by their numerical counterparts
c$CROPDMG <- ifelse((c$CROPDMGEXP =="h"| c$CROPDMGEXP =="H"),c$CROPDMG*10^2,c$CROPDMG)
c$CROPDMG <- ifelse((c$CROPDMGEXP =="k"| c$CROPDMGEXP =="K"),c$CROPDMG*10^3,c$CROPDMG)
c$CROPDMG <- ifelse((c$CROPDMGEXP =="m"| c$CROPDMGEXP =="M"),c$CROPDMG*10^6,c$CROPDMG)
c$CROPDMG <- ifelse((c$CROPDMGEXP =="b"| c$CROPDMGEXP =="B"),c$CROPDMG*10^9,c$CROPDMG)
##Aggrgating Total Crop Damage by Event Type
agg_cropdmg <- aggregate(CROPDMG~EVTYPE, c, FUN=sum, na.rm=TRUE)
#Sorting the data so that Event causing maximum Crop Damage is on Top
agg_cropdmg <- arrange(agg_cropdmg, desc(CROPDMG))
#Selecting the top 5 Events causing maximum Crop Damage
agg_cropdmg5 <- agg_cropdmg[1:5,]
#Top Five Events Causing Crop Damage
head(agg_cropdmg5)## EVTYPE CROPDMG
## 1 DROUGHT 13972566000
## 2 FLOOD 5661968450
## 3 RIVER FLOOD 5029459000
## 4 ICE STORM 5022113500
## 5 HAIL 3025954450##Data Processing to combine FATALITIES and INJURIES to get the TOTAL
#Merging the data sets for Population Damage and Crop Damage
merge_economic <- merge(agg_propdmg, agg_cropdmg, all=TRUE)
merge_economic[is.na(merge_economic)] <- 0
merge_economic <- mutate(merge_economic, TOTAL = PROPDMG + CROPDMG)
merge_economic <- arrange(merge_economic, desc(TOTAL))#Top Five Events affecting Public Health (Fatalities and Injuries)
head(merge_economic,5)## EVTYPE PROPDMG CROPDMG TOTAL
## 1 FLOOD 144657709800 5661968450 150319678250
## 2 HURRICANE/TYPHOON 69305840000 2607872800 71913712800
## 3 TORNADO 56937160480 414953110 57352113590
## 4 STORM SURGE 43323536000 5000 43323541000
## 5 HAIL 15732267220 3025954450 18758221670#Fetching Top 5 Events
merge_economic <- merge_economic[1:5,]
###Script to plot Total Fatalities and Injuries with respect to Event Type
ec <- ggplot(merge_economic, aes(x=reorder(EVTYPE, -TOTAL), y=TOTAL/10^9,FILL=EVTYPE))
ec <- ec + geom_bar(aes(fill=EVTYPE),stat="identity")
ec <- ec + geom_text(aes(label=round(TOTAL/10^9,0), hjust=0.5, vjust=-0.7),size=3.5)
ec <- ec + labs(x="\n EVENT TYPE \n")
ec <- ec + labs(y="\n TOTAL DAMAGE (in Billion $) \n")
ec <- ec + labs(title = "\n Top 5 Events With Greatest Economic Consequences \n (Considering PROPERTY and CROP Damage together)\n")
ec <- ec + coord_cartesian(ylim=c(0,165))
ec <- ec + theme(axis.text.x = element_text(angle=45, hjust=1, size=11, face = "bold"))
ec <- ec + theme(legend.position="none")
ec <- ec + theme(plot.title = element_text(size=14, face = "bold"))
#Plot for Top 5 Events With Greatest Economic Consequences
ec
RESULTS
Across the United States, most harmful types of events (in order) with respect to Population Health considering both FATALITIES and INJURIES are :
TORNADO (96,979)
EXCESSIVE HEAT (8,428)
TSTM WIND (7,461)
FLOOD (7,259)
LIGHTNING (6,046)
Across the United States, types of events (in order) that have the greatest Economic Consequences considering both Property Damage and Crop Damage are :
FLOOD (150 Billion Dollars)
HURRICANE/TYPHONE (72 Billion Dollars)
TORNADO (57 Billion Dollars)
STORM SURGE (43 Billion Dollars)
HAIL (18 Billion Dollars)