Synopsis

This report is generated from NOAA storm database generated over a period of more than 60 years (since 1950). The two issues addressed are effect on human health, and that on economy. Mortality and injuries are taken to be indicators of human health hazard; damage to property and crops are indicators of economic effect.

Data processing

The csv file (should be present in the same folder as this R markdown file) is loaded first with selected columns.

cols_to_read=c(8,23,24,25,27)
library(data.table)
d <- fread("repdata%2Fdata%2FStormData.csv",select=cols_to_read)

The dataframe is filtered for events with 1. the top 100 fatalities 2. injuries 3. property damage 4. crop damage

For each filtered dataset, the commonest event type is tabulated with the summary function, and then bar plots made.

Results

Greatest impact on human health

Fatality

df <- tail(d[order(d$FATALITIES),],100)
s_df <- summary(as.factor(df$EVTYPE))
head(s_df)
## EXCESSIVE HEAT   EXTREME HEAT    FLASH FLOOD          FLOOD           HEAT 
##             22              2              2              1              4 
##      HEAT WAVE 
##              2

Of all the events sorted by number of fatal cases, among the top 100, tornados are the commonest (58) and excessive heat the second comonext (22).

Injuries

di <- tail(d[order(d$INJURIES),],100)
s_di <- summary(as.factor(di$EVTYPE))
head(s_di)
##          BLIZZARD    EXCESSIVE HEAT             FLOOD              HEAT 
##                 1                 7                 9                 5 
##         HEAT WAVE HURRICANE/TYPHOON 
##                 1                 2

Tornado (73), flood (9) and excessive heat (7) are the top events causing injuries.

dh <- tail(d[order(d$INJURIES + d$FATALITIES),],100)
s_dh <- summary(as.factor(dh$EVTYPE))
par(mfrow=c(3,1))
barplot(head(s_dh),main="Top causes of human health hazard (fatality + injury)")
barplot(head(s_df),main="Top causes of fatality")
barplot(head(s_di),main="Top causes of injury")

Property damage

dp <- tail(d[order(d$PROPDMG),],100)
s_dp <- summary(as.factor(dp$EVTYPE))
head(s_dp)
##      BLIZZARD COASTAL FLOOD   FLASH FLOOD         FLOOD          HAIL 
##             1             1            11            29             3 
##    HEAVY SNOW 
##             1
barplot(head(s_dp),main="Top causes of property damage")

Floods, tornado, flash flood and hig hwinds are leading cause of property damege.

Crop damage

dc <- tail(d[order(d$CROPDMG),],100)
s_dc <- summary(as.factor(dc$EVTYPE))
head(s_dc)
##     DROUGHT  DUST STORM FLASH FLOOD       FLOOD        HAIL  HEAVY SNOW 
##           5           2          22          17          31           1
barplot(head(s_dc),main="Top causes of crop damage")

Hail, flash floods, floods, tornado and high winds are top causes of crop damage.