Analysis of the impact of different weather events on the local population and economy

In this report we will attempt to answer the two following questions regarding the effect of different types of storms on local populations and economies:

1. Across the United States, which types of events are most harmful with respect to population health?

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

The data to be analyzed was collected between 1950 and 2011. Earlier years will tend to have fewer data, probably due to lack of proper record keeping.

Data Processing

The following code is designed to read in the collected data and transform it into a usable form. The steps taken in the transformation are as follows:

• Create a new field called “year” which contained the year of the recorded event
• Subset the data by removing any data before 1996, as the data before this time is unreliable.
• Change the “EVTYPE” field (event type) to lower case, to facitilate easier transformations.
• Create a new field called “mainevent”.
• Assign an overarching event type to each individual recorded event and input that value into the newly created “mainevent” field.

stormdata <- read.csv("stormdata.zip")
stormdata$year <- strptime(stormdata$BGN_DATE, "%m/%d/%Y %H:%M:%S")
stormdata$year <- format(stormdata$year, "%Y")
stormdata <- stormdata[stormdata$year>1995,]
stormdata$EVTYPE <- tolower(stormdata$EVTYPE)
stormdata <- cbind(mainevent = c("other"),stormdata)
stormdata$mainevent <- as.character(stormdata$mainevent)
eventlist <- c("Astronomical Low Tide"
,"Avalanche"
,"Blizzard"
,"Flood"
,"Coastal Flood"
,"Debris Flow"
,"Dense Fog"
,"Dense Smoke"
,"Drought"
,"Dust Devil"
,"Dust Storm"
,"Excessive Heat"
,"Flash Flood"
,"Funnel Cloud"
,"Freezing Fog"
,"Hail"
,"Heat"
,"Heavy Rain"
,"Heavy Snow"
,"High Surf"
,"High Wind"
,"Ice Storm"
,"Lake-Effect Snow"
,"Lakeshore Flood"
,"Lightning"
,"Marine Hail"
,"Marine High Wind"
,"Marine Strong Wind"
,"Marine Thunderstorm Wind"
,"Rip Current"
,"Seiche"
,"Sleet"
,"Strong Wind"
,"Thunderstorm Wind"
,"Tornado"
,"Tropical Depression"
,"Tropical Storm"
,"Tsunami"
,"Volcanic Ash"
,"Waterspout"
,"Wildfire"
,"Winter Storm"
,"Winter Weather")
eventlist <- tolower(eventlist)

for(o in eventlist)
{
  for (i in grep(o,stormdata$EVTYPE))
        {
        stormdata$mainevent[i] <- o
        }

}

for (i in grep("extreme cold chill",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Extreme Cold/Wind Chill"
  }

for (i in grep("extreme wind chill",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Extreme Cold/Wind Chill"
  }

for (i in grep("storm surge",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Storm Surge/Tide"
  }

for (i in grep("storm tide",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Storm Surge/Tide"
  }

for (i in grep("hurricane",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Hurricane/Typhoon"
  }

for (i in grep("typhoon",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Hurricane/Typhoon"
  }

for (i in grep("cold chill",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Cold/Wind Chill"
  }

for (i in grep("wind chill",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Cold/Wind Chill"
  }

for (i in grep("frost",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Frost/Freeze"
  }

for (i in grep("freeze",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Frost/Freeze"
  }

for (i in grep("tstm wind",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Marine Thunderstorm Wind"
  }

for (i in grep("extreme cold",stormdata$EVTYPE))
  {
  stormdata$mainevent[i] <- "Extreme Cold/Wind Chill"
  }

Results

Across the United States, which types of events are most harmful with respect to population health?

The data collected includes number of fatalities and injuries. These are the two factors we’ll be using to determine a specific event’s impact on the health of the population.

par(mfrow=c(3,1))
plotdata <- tapply(stormdata$FATALITIES,stormdata$mainevent,sum)
plotdata2 <- tapply(stormdata$INJURIES,stormdata$mainevent,sum)
barplot(plotdata,las=2,main="Fatalities")
plot.new()
barplot(plotdata2,las=2,main="Injuries")

The charts above illustrate that for Fatalities the most harmful events are:

• Tornado
• Heat
• Flash Flood

For Injuries the most harmful events are:

• Tornado
• Heat
• Flood

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

par(mfrow=c(3,1))
plotdata <- tapply(stormdata$PROPDMG,stormdata$mainevent,sum)
plotdata2 <- tapply(stormdata$CROPDMG,stormdata$mainevent,sum)
barplot(plotdata,las=2,main="Property Damage")
plot.new()
barplot(plotdata2,las=2,main="Crop Damage")

The charts above illustrate that for Fatalities the most harmful events are:

• Marine Thunderstorm Wind
• Tornado
• Flash Flood

For Injuries the most harmful events are:

• Hail
• Flash Flood
• Flood