Effect Of Storm Events On Health And The Economy In The United States

Synopsis:

Health effects of different storm events were evaluated by taking the sum of the injuries and the fatalities for each type of event in the United States from 1050 to 2011.Tornadoes by far the storm event most harmful to human health by either assessment.Heat related events were also vert significant,especially as assessed by fatalities.Floods,lightning and thunderstorm wind also caused significant harm to the population.The economics consequences of different storm events were evaluated by taking the sum of property damage and crop damage for each type of event in the United States from 1050 to 2011.Hail caused the most property damage,but flooding,wind and lightning also caused very significant property damage.Hail caused the greatest amount of crop damage by far,but wind and flooding also caused significant crop damage.Overall the economic consequences were about eight-fold greater for property damage compared to crop damage.

Data Processing

The storm data were loaded from a csv file into a data frame. The data were cached since loading the data was time consuming.

The sums of the fatalities and injuries for each event were calculated, and subsets of the data including only events that comprised at least one percent of the totals of the fatalities and injuries, respectively, were taken to assess the significant health-related events. The totals for these events were summarized in a table and in bar plots for fatalities and for injuries. Factors for the event types were transformed to characters when combining the vectors of event types for fatalities and injuries.

Similarly, the sums of property damage and crop damage for each event were calculated, and subsets of the data including only events that comprised at least one percent of the totals of the property and crop damage, respectively, were taken to assess the economically significant events. The totals for these events were summarized in a table and in bar plots for property damage and crop damage. Factors for the event types were transformed to characters when combining the vectors of event types for property and crop damage.

findFile <- list.files(pattern = "bz2",full.names = TRUE)
stormData <- read.csv(bzfile(findFile))

# Taking sum of fatalities across event types
fatalities <- aggregate(FATALITIES ~ EVTYPE,stormData,sum)
fatalities <- fatalities[order(fatalities$FATALITIES,decreasing = TRUE),]

topFatalities <- fatalities[fatalities$FATALITIES > 0.01 * sum(fatalities$FATALITIES),]
head(topFatalities,2)
##             EVTYPE FATALITIES
## 834        TORNADO       5633
## 130 EXCESSIVE HEAT       1903
# Taking sum of injuries
injuries <- aggregate(INJURIES ~ EVTYPE,stormData,sum)
injuries <- injuries[order(injuries$INJURIES,decreasing = TRUE),]
head(injuries,2)
##        EVTYPE INJURIES
## 834   TORNADO    91346
## 856 TSTM WIND     6957
topInjuries <- injuries[injuries$INJURIES > 0.01 * sum(injuries$INJURIES),]
head(topInjuries,2)
##        EVTYPE INJURIES
## 834   TORNADO    91346
## 856 TSTM WIND     6957
topHealth <- unique(c(as.character(topFatalities$EVTYPE),as.character(topInjuries$EVTYPE)))

# Combining fatalities and injuries for the combined event type in table

topEventHealth <- cbind(sapply(topHealth,function(x) fatalities[fatalities$EVTYPE == x,"FATALITIES"]),
                     sapply(topHealth,function(x) injuries[injuries$EVTYPE==x,"INJURIES"]))

colnames(topEventHealth) <- c("fatalities","injuries")

# taking sum of property damage

propertyDamage <- aggregate(PROPDMG~EVTYPE,stormData,sum)
propertyDamage <- propertyDamage[order(propertyDamage$PROPDMG,decreasing = TRUE),]

topDamageProp <- propertyDamage[propertyDamage$PROPDMG > 0.01 * sum(propertyDamage$PROPDMG),]

# taking sum of crop damage
cropdamage <- aggregate(CROPDMG~EVTYPE,stormData,sum)
cropdamage <- cropdamage[order(cropdamage$CROPDMG,decreasing = TRUE),]

topCropDamage <- cropdamage[cropdamage$CROPDMG > 0.01 * sum(cropdamage$CROPDMG),]

topPCdamage <- unique(c(as.character(topDamageProp$EVTYPE),as.character(topCropDamage$EVTYPE)))

# combining property and crop damage for the combined event types in table

topEventDamage <- cbind(sapply(topPCdamage,function(x) propertyDamage[  propertyDamage$EVTYPE==x,"PROPDMG"]),sapply(topPCdamage,function(x) cropdamage[cropdamage$EVTYPE==x,"CROPDMG"]))

colnames(topEventDamage) <- c("property.damage","crop.damage")

Results

US Storm Events Most Harmful To Health

Table 1 Total fatalities and injuries for storm events that caused at least 1% of the either the fatalities or injuries

head(topEventHealth,13)
##                fatalities injuries
## TORNADO              5633    91346
## EXCESSIVE HEAT       1903     6525
## FLASH FLOOD           978     1777
## HEAT                  937     2100
## LIGHTNING             816     5230
## TSTM WIND             504     6957
## FLOOD                 470     6789
## RIP CURRENT           368      232
## HIGH WIND             248     1137
## AVALANCHE             224      170
## WINTER STORM          206     1321
## RIP CURRENTS          204      297
## HEAT WAVE             172      309

Figure-1 Health effects of the storm events that caused at least 1% of the fatalities

barplot(topEventHealth[,"fatalities"],col = "orange",main = "Us Storm Events Causing The Most Fatalities",ylab = "Total Fatalities Per Event Type",las=2,cex.names = 0.55,cex.axis = 0.70)

Figure-2 Health effects of the storm events that caused at least 1% of the injuries

barplot(topEventHealth[,"injuries"],las=2,col = "purple",main = "Us Storm Events Causing The Most Injuries",ylab = "Total Injuries Per Event Type",cex.names = 0.55,cex.axis = 0.70)

print(" ")
## [1] " "

US Storm Events with the Greatest Economic Consequences

Table-2 Total property damage and crop damage for storm events that caused at least 1% of the either the property or crop damage.

head(topEventDamage,13)
##                    property.damage crop.damage
## TORNADO                 3212258.16   100018.52
## FLASH FLOOD             1420124.59   179200.46
## TSTM WIND               1335965.61   109202.60
## FLOOD                    899938.48   168037.88
## THUNDERSTORM WIND        876844.17    66791.45
## HAIL                     688693.38   579596.28
## LIGHTNING                603351.78     3580.61
## THUNDERSTORM WINDS       446293.18    18684.93
## HIGH WIND                324731.56    17283.21
## WINTER STORM             132720.59     1978.99
## HEAVY SNOW               122251.99     2165.72
## DROUGHT                    4099.05    33898.62

Figure-3 - The relative relative economic consequences of the storm events that caused at least 1% of the either the property damage or the crop damage.

par(mfrow=c(1,2),mar=c(15,4,3,2),mgp=c(3,1,0),cex=0.8)

barplot(round(topEventDamage[,"property.damage"] / 1000,0),main = "US Storm Property Damage",ylab = "Total Property Damage",las=2,col = "wheat")

barplot(round(topEventDamage[,"crop.damage"] / 1000,0),main = "US Storm Crop Damage",ylab = "Total Crop Damage",las=2,col = "pink")

print(" ")
## [1] " "