This analysis downloads data from the US National Oceanic and Atmospheric Administration's (NOAA) storm database to determine which severe whether events have the biggest human and economic effects in the US.
Results demonstrate that
The data for this assignment was downloaded directly from the NOAA website.
setwd("C:/Users/Owner/Documents/DataScienceCertification/RepoResearch/peer2")
myURL<-"http://d396qusza40orc.cloudfront.net/repdata%2Fdata%2FStormData.csv.bz2"
download.file(myURL,"StormData.csv.bz2")
df1<-read.csv(bzfile("StormData.csv.bz2"), stringsAsFactors=FALSE)
Next, a subset of the data frame relevant for the analysis is taken. The subset includes
1. Only the columns required for the analysis, and
2. Only rows for which there is non-zero human and/or economic impact
# Select Only Relevant Columns for Analysis
df2<-df1[,c("EVTYPE","FATALITIES","INJURIES","PROPDMG","PROPDMGEXP","CROPDMG","CROPDMGEXP")]
# Select Only Rows with Non Zero Human or Economic Costs
df3<-df2[rowSums(df2[,c("FATALITIES","INJURIES","PROPDMG","CROPDMG")])>0,]
Some of the events listed in the database are coded with various lowercase and uppercase formats, but represent the same weather events. Therefore, event type was converted to all uppercase to consolidate these cases.
# Format All Event Types in Uppercase
df3$EVTYPE.Upper<-toupper(df3$EVTYPE)
In order to estimate economic damage, it is necessary to transform the economic measures. For both property damage and crop damage, there is a column of numeric values and a column of multipliers. The column of multipliers indicates the magnitude of the numeric values (H=Hundreds, K=Thousands, M=Millions, and B=Billions). As with event type, it is necessary to convert the multipliers to consistent values (i.e. all uppercase). Then a function is applied to transform the combination of the numerical column and the column of multipliers into the actual economic costs associated with property damage and crop damage.
# Format All DMGEXP in Uppercase
df3$PROPDMGEXP<-toupper(df3$PROPDMGEXP)
df3$CROPDMGEXP<-toupper(df3$CROPDMGEXP)
# Transormation Function
convertMe<-function(x){if(x=="H"){
100
} else if(x=="K"){
1000
} else if(x=="M"){
1000000
} else if(x=="B"){
1000000000
} else {
1
}
}
df3$P.DMG<-df3$PROPDMG*sapply(df3$PROPDMGEXP,convertMe)
df3$C.DMG<-df3$CROPDMG*sapply(df3$CROPDMGEXP,convertMe)
Once the economic variables have been transformed, the data is once again subset to obtain only the data required for the analysis.
df4<-df3[,c("EVTYPE.Upper","FATALITIES","INJURIES","P.DMG","C.DMG")]
Now the data is summarized in two distinct data tables.
The first sums the number of fatalities and injuries associated with each type of severe weather event. Note that since we are interested in studying the most severe events, those event types with less than or equal to 10 total fatalities plus injuries are excluded.
Similarly, a table summing the economic impact on property and crops is calculated such that the total value is greater than $1,000,000.
library(reshape2)
## Warning: package 'reshape2' was built under R version 3.0.3
# Summarize Fatalities and Injuries by EVTYPE
dfFatalities<-dcast(data=df4, EVTYPE.Upper~ ., value.var="FATALITIES", sum)
names(dfFatalities)<-c("Event.Type","Fatalities")
dfInjuries<-dcast(data=df4, EVTYPE.Upper~ ., value.var="INJURIES", sum)
names(dfInjuries)<-c("Event.Type","Injuries")
dfHuman<-merge(dfFatalities,dfInjuries)
# Select Only Big Events Affecting Human Well Being
dfHuman<- dfHuman[dfHuman$Fatalities + dfHuman$Injuries > 10,]
dfHuman<- dfHuman[order(dfHuman$Fatalities,decreasing=TRUE),]
# Summarize Economic Impact by EVTYPE
dfProp<-dcast(data=df4, EVTYPE.Upper~ ., value.var="P.DMG", sum)
names(dfProp)<-c("Event.Type","Property.Damage")
dfCrop<-dcast(data=df4, EVTYPE.Upper~ ., value.var="C.DMG", sum)
names(dfCrop)<-c("Event.Type","Crop.Damage")
dfEcon<-merge(dfProp,dfCrop)
# Select Only Big Events Affecting Economic Well Being
dfEcon<- dfEcon[dfEcon$Property.Damage + dfEcon$Crop.Damage > 1000000,]
dfEcon<- dfEcon[order(dfEcon$Property.Damage,decreasing=TRUE),]
Finally, process the data to generate plots that will be included in the results section. Note: only the top 20 severe weather events are included in the plots.
# Graph Human Impact
library(ggplot2)
## Warning: package 'ggplot2' was built under R version 3.0.3
library(reshape2)
dfHumanLong<-melt(dfHuman[1:20,],id.var=c("Event.Type"))
pHuman<- ggplot(data=dfHumanLong,aes(x=Event.Type,y=value))
pHuman<- pHuman + geom_bar(stat="identity")
pHuman<- pHuman + facet_wrap(~variable, scales="free")
pHuman<- pHuman + theme(axis.text.x=element_text(angle=90,hjust=1,vjust=0.5))
pHuman<- pHuman + ylab("Number of People") + ggtitle("Human Impact From Severe Weather Events in the US")
# Graph Economic Impact
library(ggplot2)
library(reshape2)
dfEconLong<-melt(dfEcon[1:20,],id.var=c("Event.Type"))
pEcon<- ggplot(data=dfEconLong,aes(x=Event.Type,y=value))
pEcon<- pEcon + geom_bar(stat="identity")
pEcon<- pEcon + facet_wrap(~variable, scales="free")
pEcon<- pEcon + theme(axis.text.x=element_text(angle=90,hjust=1,vjust=0.5))
pEcon<- pEcon + ylab("US Dollars ($)") + ggtitle("Economic Impact From Severe Weather Events in the US")
The table and plot (below) summarize the total number of fatalities and injuries associated with each severe weather event. While some of the severe weather events could certainly be consolidated, it is clear that far more deaths and injuries occur with tornadoes than any other weather event.
dfHuman
## Event.Type Fatalities Injuries
## 371 TORNADO 5633 91346
## 54 EXCESSIVE HEAT 1903 6525
## 65 FLASH FLOOD 978 1777
## 133 HEAT 937 2100
## 233 LIGHTNING 816 5230
## 386 TSTM WIND 504 6957
## 78 FLOOD 470 6789
## 276 RIP CURRENT 368 232
## 179 HIGH WIND 248 1137
## 11 AVALANCHE 224 170
## 441 WINTER STORM 206 1321
## 277 RIP CURRENTS 204 297
## 134 HEAT WAVE 172 379
## 59 EXTREME COLD 162 231
## 329 THUNDERSTORM WIND 133 1488
## 151 HEAVY SNOW 127 1021
## 60 EXTREME COLD/WIND CHILL 125 24
## 174 HIGH SURF 104 156
## 319 STRONG WIND 103 280
## 14 BLIZZARD 101 805
## 140 HEAVY RAIN 98 251
## 61 EXTREME HEAT 96 155
## 35 COLD/WIND CHILL 95 12
## 216 ICE STORM 89 1975
## 433 WILDFIRE 75 911
## 203 HURRICANE/TYPHOON 64 1275
## 346 THUNDERSTORM WINDS 64 908
## 90 FOG 62 734
## 194 HURRICANE 61 46
## 381 TROPICAL STORM 58 340
## 169 HEAVY SURF/HIGH SURF 42 48
## 28 COLD 38 48
## 224 LANDSLIDE 38 52
## 187 HIGH WINDS 35 302
## 403 TSUNAMI 33 129
## 444 WINTER WEATHER 33 398
## 409 UNSEASONABLY WARM AND DRY 29 0
## 419 URBAN/SML STREAM FLD 28 79
## 446 WINTER WEATHER/MIX 28 72
## 377 TORNADOES, TSTM WIND, HAIL 25 0
## 435 WIND 23 86
## 50 DUST STORM 22 440
## 74 FLASH FLOODING 19 8
## 40 DENSE FOG 18 342
## 63 EXTREME WINDCHILL 17 5
## 82 FLOOD/FLASH FLOOD 17 15
## 275 RECORD/EXCESSIVE HEAT 17 0
## 116 HAIL 15 1361
## 30 COLD AND SNOW 14 0
## 72 FLASH FLOOD/FLOOD 14 0
## 249 MARINE STRONG WIND 14 22
## 317 STORM SURGE 13 38
## 431 WILD/FOREST FIRE 12 545
## 318 STORM SURGE/TIDE 11 5
## 408 UNSEASONABLY WARM 11 17
## 250 MARINE THUNDERSTORM WIND 10 26
## 443 WINTER STORMS 10 17
## 251 MARINE TSTM WIND 9 8
## 166 HEAVY SURF 8 40
## 284 ROUGH SEAS 8 5
## 320 STRONG WINDS 8 21
## 384 TROPICAL STORM GORDON 8 43
## 96 FREEZING RAIN 7 23
## 104 GLAZE 7 216
## 248 MARINE MISHAP 7 5
## 208 ICE 6 137
## 173 HIGH SEAS 5 8
## 219 ICY ROADS 5 31
## 296 SNOW 5 31
## 400 TSTM WIND/HAIL 5 95
## 115 GUSTY WINDS 4 11
## 135 HEAT WAVE DROUGHT 4 15
## 46 DRY MICROBURST 3 28
## 182 HIGH WIND AND SEAS 3 20
## 422 WATERSPOUT 3 29
## 427 WATERSPOUT/TORNADO 3 42
## 430 WILD FIRES 3 150
## 17 BLOWING SNOW 2 14
## 48 DUST DEVIL 2 43
## 55 EXCESSIVE RAINFALL 2 21
## 94 FREEZING DRIZZLE 2 15
## 255 MIXED PRECIP 2 26
## 272 RECORD HEAT 2 50
## 302 SNOW SQUALL 2 35
## 13 BLACK ICE 1 24
## 325 THUNDERSTORM 1 12
## 442 WINTER STORM HIGH WINDS 1 15
## 447 WINTRY MIX 1 77
## 106 GLAZE/ICE STORM 0 15
## 310 SNOW/HIGH WINDS 0 36
## 365 THUNDERSTORMW 0 27
## 374 TORNADO F2 0 16
## 445 WINTER WEATHER MIX 0 68
pHuman
The table and plot (below) summarize the total cost of property damage and crop damage associated with each severe weather event. While some of the severe weather events could certainly be consolidated, it is clear that flooding and events that lead to flooding have the largest economic impact in the US, especially for property damage. However, note that ice storms and hail also have a large impact on crop damage.
dfEcon
## Event.Type Property.Damage Crop.Damage
## 78 FLOOD 1.447e+11 5.662e+09
## 203 HURRICANE/TYPHOON 6.931e+10 2.608e+09
## 371 TORNADO 5.694e+10 4.150e+08
## 317 STORM SURGE 4.332e+10 5.000e+03
## 65 FLASH FLOOD 1.614e+10 1.421e+09
## 116 HAIL 1.573e+10 3.026e+09
## 194 HURRICANE 1.187e+10 2.742e+09
## 381 TROPICAL STORM 7.704e+09 6.783e+08
## 441 WINTER STORM 6.688e+09 2.694e+07
## 179 HIGH WIND 5.270e+09 6.386e+08
## 280 RIVER FLOOD 5.119e+09 5.029e+09
## 433 WILDFIRE 4.765e+09 2.955e+08
## 318 STORM SURGE/TIDE 4.641e+09 8.500e+05
## 386 TSTM WIND 4.485e+09 5.540e+08
## 216 ICE STORM 3.945e+09 5.022e+09
## 329 THUNDERSTORM WIND 3.483e+09 4.148e+08
## 201 HURRICANE OPAL 3.173e+09 1.900e+07
## 431 WILD/FOREST FIRE 3.002e+09 1.068e+08
## 144 HEAVY RAIN/SEVERE WEATHER 2.500e+09 0.000e+00
## 346 THUNDERSTORM WINDS 1.736e+09 1.907e+08
## 377 TORNADOES, TSTM WIND, HAIL 1.600e+09 2.500e+06
## 288 SEVERE THUNDERSTORM 1.205e+09 2.000e+05
## 43 DROUGHT 1.046e+09 1.397e+10
## 151 HEAVY SNOW 9.326e+08 1.347e+08
## 233 LIGHTNING 9.287e+08 1.209e+07
## 140 HEAVY RAIN 6.942e+08 7.334e+08
## 14 BLIZZARD 6.592e+08 1.121e+08
## 430 WILD FIRES 6.241e+08 0.000e+00
## 187 HIGH WINDS 6.083e+08 4.072e+07
## 405 TYPHOON 6.002e+08 8.250e+05
## 224 LANDSLIDE 3.246e+08 2.002e+07
## 74 FLASH FLOODING 3.078e+08 1.512e+07
## 72 FLASH FLOOD/FLOOD 2.725e+08 5.550e+05
## 22 COASTAL FLOOD 2.596e+08 0.000e+00
## 198 HURRICANE ERIN 2.581e+08 1.360e+08
## 130 HAILSTORM 2.410e+08 0.000e+00
## 319 STRONG WIND 1.753e+08 6.495e+07
## 82 FLOOD/FLASH FLOOD 1.740e+08 9.503e+07
## 403 TSUNAMI 1.441e+08 2.000e+04
## 23 COASTAL FLOODING 1.330e+08 5.600e+04
## 281 RIVER FLOODING 1.163e+08 2.802e+07
## 191 HIGH WINDS/COLD 1.105e+08 7.000e+06
## 87 FLOODING 1.083e+08 8.856e+06
## 244 MAJOR FLOOD 1.050e+08 0.000e+00
## 434 WILDFIRES 1.005e+08 5.000e+05
## 202 HURRICANE OPAL/HIGH WINDS 1.000e+08 1.000e+07
## 174 HIGH SURF 8.995e+07 0.000e+00
## 59 EXTREME COLD 6.774e+07 1.313e+09
## 442 WINTER STORM HIGH WINDS 6.000e+07 5.000e+06
## 419 URBAN/SML STREAM FLD 5.831e+07 8.488e+06
## 271 RECORD COLD 5.600e+07 0.000e+00
## 427 WATERSPOUT/TORNADO 5.111e+07 0.000e+00
## 197 HURRICANE EMILY 5.000e+07 0.000e+00
## 400 TSTM WIND/HAIL 4.434e+07 6.470e+07
## 220 LAKE-EFFECT SNOW 4.011e+07 0.000e+00
## 444 WINTER WEATHER 2.087e+07 1.500e+07
## 24 COASTAL FLOODING/EROSION 2.003e+07 0.000e+00
## 413 URBAN FLOOD 1.829e+07 9.235e+05
## 53 EROSION/CSTL FLOOD 1.620e+07 0.000e+00
## 20 COASTAL FLOODING/EROSION 1.500e+07 0.000e+00
## 296 SNOW 1.483e+07 1.000e+04
## 142 HEAVY RAIN/HIGH SURF 1.350e+07 1.500e+06
## 90 FOG 1.316e+07 0.000e+00
## 208 ICE 1.265e+07 0.000e+00
## 147 HEAVY RAINS 1.229e+07 6.050e+07
## 101 FROST/FREEZE 1.048e+07 1.094e+09
## 134 HEAT WAVE 1.046e+07 5.550e+06
## 79 FLOOD & HEAVY RAIN 1.000e+07 0.000e+00
## 169 HEAVY SURF/HIGH SURF 9.870e+06 0.000e+00
## 40 DENSE FOG 9.674e+06 0.000e+00
## 8 ASTRONOMICAL HIGH TIDE 9.425e+06 0.000e+00
## 422 WATERSPOUT 9.354e+06 0.000e+00
## 77 FLASH FLOODS 8.728e+06 0.000e+00
## 435 WIND 8.694e+06 3.000e+05
## 60 EXTREME COLD/WIND CHILL 8.648e+06 5.000e+04
## 96 FREEZING RAIN 8.146e+06 0.000e+00
## 3 TSTM WIND 8.100e+06 0.000e+00
## 39 DAMAGING FREEZE 8.000e+06 2.962e+08
## 54 EXCESSIVE HEAT 7.754e+06 4.924e+08
## 223 LAKESHORE FLOOD 7.540e+06 0.000e+00
## 188 HIGH WINDS HEAVY RAINS 7.500e+06 1.000e+04
## 46 DRY MICROBURST 6.733e+06 1.500e+04
## 446 WINTER WEATHER/MIX 6.372e+06 0.000e+00
## 89 FLOODS 6.000e+06 5.000e+04
## 414 URBAN FLOODING 5.606e+06 3.665e+05
## 50 DUST STORM 5.549e+06 3.100e+06
## 213 ICE JAM FLOODING 5.516e+06 5.000e+06
## 251 MARINE TSTM WIND 5.421e+06 0.000e+00
## 266 RAIN 5.300e+06 7.500e+05
## 315 SNOWMELT FLOODING 5.050e+06 0.000e+00
## 92 FOREST FIRES 5.000e+06 5.000e+05
## 143 HEAVY RAIN/LIGHTNING 5.000e+06 0.000e+00
## 158 HEAVY SNOW/BLIZZARD/AVALANCHE 5.000e+06 0.000e+00
## 159 HEAVY SNOW/FREEZING RAIN 5.000e+06 0.000e+00
## 165 HEAVY SNOWPACK 5.000e+06 0.000e+00
## 190 HIGH WINDS/COASTAL FLOOD 5.000e+06 0.000e+00
## 209 ICE AND SNOW 5.000e+06 0.000e+00
## 237 LIGHTNING FIRE 5.000e+06 0.000e+00
## 382 TROPICAL STORM ALBERTO 5.000e+06 0.000e+00
## 148 HEAVY RAINS/FLOODING 4.860e+06 3.530e+05
## 385 TROPICAL STORM JERRY 4.600e+06 1.600e+07
## 11 AVALANCHE 3.722e+06 0.000e+00
## 218 ICE/STRONG WINDS 3.500e+06 0.000e+00
## 230 LIGHT SNOW 2.513e+06 0.000e+00
## 320 STRONG WINDS 2.415e+06 5.000e+06
## 373 TORNADO F1 2.370e+06 0.000e+00
## 95 FREEZING FOG 2.182e+06 0.000e+00
## 290 SEVERE THUNDERSTORMS 2.000e+06 1.700e+07
## 35 COLD/WIND CHILL 1.990e+06 6.000e+05
## 56 EXCESSIVE SNOW 1.935e+06 0.000e+00
## 360 THUNDERSTORM WINDSS 1.836e+06 5.955e+04
## 133 HEAT 1.797e+06 4.015e+08
## 75 FLASH FLOODING/FLOOD 1.750e+06 1.750e+05
## 380 TROPICAL DEPRESSION 1.737e+06 0.000e+00
## 154 HEAVY SNOW AND STRONG WINDS 1.700e+06 0.000e+00
## 374 TORNADO F2 1.600e+06 0.000e+00
## 160 HEAVY SNOW/HIGH WINDS & FLOOD 1.500e+06 2.000e+04
## 166 HEAVY SURF 1.440e+06 0.000e+00
## 313 SNOW/SLEET 1.350e+06 0.000e+00
## 138 HEAVY MIX 1.305e+06 0.000e+00
## 301 SNOW FREEZING RAIN 1.305e+06 0.000e+00
## 247 MARINE HIGH WIND 1.297e+06 0.000e+00
## 115 GUSTY WINDS 1.281e+06 2.000e+05
## 260 MUDSLIDE 1.275e+06 0.000e+00
## 279 RIVER AND STREAM FLOOD 1.200e+06 0.000e+00
## 308 SNOW/FREEZING RAIN 1.200e+06 0.000e+00
## 183 HIGH WIND DAMAGE 1.100e+06 0.000e+00
## 38 DAM BREAK 1.002e+06 0.000e+00
## 325 THUNDERSTORM 9.206e+05 1.000e+06
## 63 EXTREME WINDCHILL 7.550e+05 1.700e+07
## 440 WINDS 6.460e+05 5.050e+05
## 93 FREEZE 2.050e+05 4.567e+08
## 289 SEVERE THUNDERSTORM WINDS 1.750e+05 2.900e+07
## 61 EXTREME HEAT 1.150e+05 5.000e+06
## 294 SMALL HAIL 7.000e+04 2.079e+07
## 265 OTHER 5.550e+04 1.034e+06
## 100 FROST 1.500e+04 6.600e+07
## 6 AGRICULTURAL FREEZE 0.000e+00 2.882e+07
## 31 COLD AND WET CONDITIONS 0.000e+00 6.600e+07
## 37 COOL AND WET 0.000e+00 5.000e+06
## 52 EARLY FROST 0.000e+00 4.200e+07
## 57 EXCESSIVE WETNESS 0.000e+00 1.420e+08
## 85 FLOOD/RAIN/WINDS 0.000e+00 1.128e+08
## 131 HARD FREEZE 0.000e+00 1.310e+07
## 406 UNSEASONABLE COLD 0.000e+00 5.100e+06
## 407 UNSEASONABLY COLD 0.000e+00 2.504e+07
## 410 UNSEASONAL RAIN 0.000e+00 1.000e+07
pEcon
