Severe Weather Events Health and Economic Impacts in the US (1996-2016)

Synopsis

Storms and other sever weather events can cause both public and eonomic 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 involved exploring the U.S. National Oceanic and Atmospheric Adminnistration’s (NOAA) storm database to determine the severe weather events that had the greatest health and economic impacts.

library(ggplot2)
library(knitr)

Data Processing

The data was downloaded from the NOAA database and loaded into R as a data frame.

##URL <- "https://d396qusza40orc.cloudfront.net/repdata%2Fdata%2FStrormData.csv.bz2"
destfile <- "C:/Users/Dave/Documents/DataScienceSpecialization/ReproducibleResearch/Project02/Working/StormData.csv"
##download.file(URL,destfile)
StormDataRaw <- read.csv((destfile), strip.white = TRUE)

On examining the data it was determined that only select columns were needed for the analysis and that data prior to 1996 was incomplete. In addition the Event Type data column was formated to uppper case as the case of the events was not always consistent and could cause problems with summarizing the data.

## Convert BGN_DATE to date type
 StormDataRaw$BGN_DATE <- as.Date(StormDataRaw$BGN_DATE, format = "%m/%d/%Y")

## Subset the data on health and damage relevant columns
StormData <- subset(StormDataRaw, select = c("EVTYPE", "FATALITIES", "INJURIES", "PROPDMG", "PROPDMGEXP", "CROPDMG", "CROPDMGEXP", "BGN_DATE", "REMARKS"))
StromData <- subset(StormData, StormData$BGN_DATE > as.Date("1995-12-31"))

## Convert EVTYPE to upper case
StormData$EVTYPE <- toupper(StormData$EVTYPE)

Property and crop damage columns recorded economic data in two columns. PROPDMGEXP and CROPDMGEXP recorded the modifier of the damage columns as characters. It was necessary to convert the modifier to numeric and then calculate the economic damage in U.S. dollars.

## Calculate DMG in US dollars
format <- function(x){
        x <- as.character(x)
        ifelse (x == "H", as.numeric(100),
        ifelse (x == "K", as.numeric(1000),
        ifelse (x == "M", as.numeric(1000000),
        ifelse (x == "B", as.numeric(1000000000), 0))))
}
## Calculate property DMG in US dollars
StormData$PROPDMGEXP <- toupper(StormData$PROPDMGEXP)
StormData$PROPDMGEXP <- format(StormData$PROPDMGEXP)
StormData$PDMGUSD <- as.numeric(StormData$PROPDMG*StormData$PROPDMGEXP)

## Calculate crop DMG in US dollars
StormData$CROPDMGEXP <- toupper(StormData$CROPDMGEXP)
StormData$CROPDMGEXP <- format(StormData$CROPDMGEXP)
StormData$CDMGUSD <- as.numeric(StormData$CROPDMG*StormData$CROPDMGEXP)

To prepare the data for plotting, all relevant datasets were summarized by Event Type and records with no impact recorded were excluded.

## Sum of Health and DMG types
PropDMG <- aggregate(PDMGUSD ~ EVTYPE, StormData, FUN = sum)
CropDMG <- aggregate(CDMGUSD ~ EVTYPE, StormData, FUN = sum)
Fatal <- aggregate(FATALITIES ~ EVTYPE, StormData, FUN= sum)
Injury <- aggregate(INJURIES ~ EVTYPE, StormData, FUN = sum)

For the purposes of this analysis, only the top ten events causing the greatest health and economic impact will be used.

PropDMG <- PropDMG[(PropDMG$PDMGUSD > 0),]
CropDMG <- CropDMG[(CropDMG$CDMGUSD > 0),]
Fatal <- Fatal[(Fatal$FATALITIES > 0),]
Injury <- Injury[(Injury$INJURIES > 0),]

PropDMGsorted <- PropDMG[order(PropDMG$PDMGUSD, decreasing = TRUE),]
CropDMGsorted <- CropDMG[order(CropDMG$CDMGUSD, decreasing = TRUE),]
Fatalsorted <- Fatal[order(Fatal$FATALITIES, decreasing = TRUE),]
Injurysorted <- Injury[order(Injury$INJURIES, decreasing = TRUE),]

An R code snippet was taken from the online R Cookbook and used to create a function that plots multiple plots in one panel.

## From R Cookbook (http://www.cookbook-r.com/Graphs/Multiple_graphs_on_one_page_(ggplot2)/)

# Multiple plot function
#
# ggplot objects can be passed in ..., or to plotlist (as a list of ggplot objects)
# - cols:   Number of columns in layout
# - layout: A matrix specifying the layout. If present, 'cols' is ignored.
#
# If the layout is something like matrix(c(1,2,3,3), nrow=2, byrow=TRUE),
# then plot 1 will go in the upper left, 2 will go in the upper right, and
# 3 will go all the way across the bottom.
#
multiplot <- function(..., plotlist=NULL, file, cols=1, layout=NULL) {
  library(grid)

  # Make a list from the ... arguments and plotlist
  plots <- c(list(...), plotlist)

  numPlots = length(plots)

  # If layout is NULL, then use 'cols' to determine layout
  if (is.null(layout)) {
    # Make the panel
    # ncol: Number of columns of plots
    # nrow: Number of rows needed, calculated from # of cols
    layout <- matrix(seq(1, cols * ceiling(numPlots/cols)),
                    ncol = cols, nrow = ceiling(numPlots/cols))
  }

 if (numPlots==1) {
    print(plots[[1]])

  } else {
    # Set up the page
    grid.newpage()
    pushViewport(viewport(layout = grid.layout(nrow(layout), ncol(layout))))

    # Make each plot, in the correct location
    for (i in 1:numPlots) {
      # Get the i,j matrix positions of the regions that contain this subplot
      matchidx <- as.data.frame(which(layout == i, arr.ind = TRUE))

      print(plots[[i]], vp = viewport(layout.pos.row = matchidx$row,
                                      layout.pos.col = matchidx$col))
    }
  }
}

Results

Health Impact Analysis It was decided to only plot the top ten event types in order of decreasing impacts to health (Injuries and Fatalities). Tornadoes as an event had the greatest number of injuries and fatalities. For the other event types althought there impact on injuries and fatalities differed, it was found that generally the same events had the greatest impacts.

pInjury <- ggplot(data = head(Injurysorted,10), aes(x=reorder(EVTYPE, INJURIES), y = INJURIES)) +
geom_bar(fill="blue", stat="identity") + coord_flip() +
ylab("Total number of injuries") + xlab("Event type") +
ggtitle("Injuries") + theme(legend.position="none")

pFatal <- ggplot(data = head(Fatalsorted,10), aes(x=reorder(EVTYPE, FATALITIES), y = FATALITIES)) +
geom_bar(fill = "red", stat="identity") + coord_flip() +
ylab("Total number of fatalities") + xlab("Event Type") + ggtitle("Fatalities") + theme(legend.position="none")

multiplot(pInjury, pFatal, cols = 2)

Economic Impact Analysis It was decided to only plot the top ten event types in order of decreasing economic damage (Property and Crop). The greatest amount of property damage was from flooding and perhaps not suprisingly, the greatest amount of crop damage was from droughts.

pPropDMG <- ggplot(data=head(PropDMGsorted,10), aes(x=reorder(EVTYPE, PDMGUSD), y = PDMGUSD)) +
geom_bar(fill = "green", stat="identity") + coord_flip() + ylab("Damage (USD)") + xlab("Event Type") + ggtitle("Property damage (USD)") + theme(legend.position="none")

pCropDMG <- ggplot(data=head(CropDMGsorted,10), aes(x=reorder(EVTYPE,CDMGUSD), y = CDMGUSD)) + geom_bar(fill="orange", stat="identity") + coord_flip() +
ylab("Damage (USD)") + xlab("Event Type") + ggtitle("Crop damage (USD)") + theme(legend.position="none")

multiplot(pPropDMG, pCropDMG, cols = 2)

Reproducible Research - Project 02

David Junor

August 30, 2016

Severe Weather Events Health and Economic Impacts in the US (1996-2016)

Synopsis

Data Processing

Results