Synopsis

In terms of health related damage, tornado is the most harmful events with respect to the number of fatalities and the number of events. Excessive heat, flash flood, heat and lightning are also very harmful event as to fatalities. TSTM wind is very freuquent event which cause many fatalities. In terms of economic consequence, tornado is the most impactful event to the property damages. TSTM wind is also very harmful not only on the level of property damage and also the massive number of the events. Flash flood, flood and thunderstrom wind is also the harmful event to the property damages. Hail is very frequent, which causes large property damages.

Data Processing

storm_original <- read.csv("repdata_data_StormData.csv", header=TRUE)

Calculating the aggregation of fatalities, injuries and propaty domages. Then, binding all aggregation into agg_total

agg_fatalities <- storm_original %>% 
  group_by(EVTYPE) %>% 
  summarize(number_fatalities = sum(FATALITIES))

agg_injuries <- storm_original %>% 
  group_by(EVTYPE) %>% 
  summarize(number_injuries = sum(INJURIES))

agg_propdmg <- storm_original %>% 
  group_by(EVTYPE) %>% 
  summarize(propdmg = sum(PROPDMG))

agg_total <- left_join(agg_fatalities, agg_injuries, by="EVTYPE")
agg_total <- left_join(agg_total, agg_propdmg, by="EVTYPE")

summary of the aggregations of fatalities, injuries and property damages. Then, plot the relation of each items.

Accoring to the plot of the correlation of each items, a small number of event types caused massive damage in terms of health and economics. The government needs to prepare the most disastrous event. So I abstruct the top 20 of event types in terms of fatalities, injuries and property damages.

summary(agg_total)

##     EVTYPE          number_fatalities number_injuries      propdmg       
##  Length:985         Min.   :   0.00   Min.   :    0.0   Min.   :      0  
##  Class :character   1st Qu.:   0.00   1st Qu.:    0.0   1st Qu.:      0  
##  Mode  :character   Median :   0.00   Median :    0.0   Median :      0  
##                     Mean   :  15.38   Mean   :  142.7   Mean   :  11050  
##                     3rd Qu.:   0.00   3rd Qu.:    0.0   3rd Qu.:     35  
##                     Max.   :5633.00   Max.   :91346.0   Max.   :3212258

plot(agg_total)

top20_fatalities <- agg_total %>% 
  arrange(desc(number_fatalities)) %>% 
  head(20)

top20_injuries <- agg_total %>% 
  arrange(desc(number_injuries)) %>% 
  head(20)

top20_propdmg <- agg_total %>% 
  arrange(desc(propdmg)) %>% 
  head(20)

Calculating the number of event occurs

count_event <- storm_original %>% 
  group_by(EVTYPE) %>% 
  summarize(count=n())

Results

which types of events are most harmful with respect to population health?

Showing the table which present the top 20 event type in terms of fatalities and injuries. This table show 28 events becase some of event does not contain any number. In addition, the table includes the number of event that occured. According to this table, tornado, excessive heat, flash flooding, heat, lighting, tstm wind and flood are the most harmful with respect to population health.

top20_fatalities_select <- top20_fatalities %>% 
  select(EVTYPE, number_fatalities)
top20_injuries_select <- top20_injuries %>% 
  select(EVTYPE, number_injuries)
agg_health <- full_join(top20_fatalities_select, top20_injuries_select, by="EVTYPE")
agg_health <- left_join(agg_health, count_event, by="EVTYPE")
gt(agg_health)

EVTYPE	number_fatalities	number_injuries	count
TORNADO	5633	91346	60652
EXCESSIVE HEAT	1903	6525	1678
FLASH FLOOD	978	1777	54277
HEAT	937	2100	767
LIGHTNING	816	5230	15754
TSTM WIND	504	6957	219940
FLOOD	470	6789	25326
RIP CURRENT	368	NA	470
HIGH WIND	248	1137	20212
AVALANCHE	224	NA	386
WINTER STORM	206	1321	11433
RIP CURRENTS	204	NA	304
HEAT WAVE	172	NA	74
EXTREME COLD	160	NA	655
THUNDERSTORM WIND	133	1488	82563
HEAVY SNOW	127	1021	15708
EXTREME COLD/WIND CHILL	125	NA	1002
STRONG WIND	103	NA	3566
BLIZZARD	101	805	2719
HIGH SURF	101	NA	725
ICE STORM	NA	1975	2006
HAIL	NA	1361	288661
HURRICANE/TYPHOON	NA	1275	88
WILDFIRE	NA	911	2761
THUNDERSTORM WINDS	NA	908	20843
FOG	NA	734	538
WILD/FOREST FIRE	NA	545	1457
DUST STORM	NA	440	427

Shoing the plot indicating which type are most harmful and frequent in terms of health. X axis indicate the logarizm of the number of fatalities. Y axis indicates the type of events. The size of point indicates how frequent the event occurs.

agg_health %>% 
  drop_na(number_fatalities) %>% 
  ggplot(aes(x=log(number_fatalities), y=EVTYPE, size=count, color=count)) +
  geom_point() +
  theme_bw()

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

Showing the table which present the top 20 event type in terms of property damages. According to this table, tornado, flash flood, tstm wind, flood, thunderstorm wind and hail are the most harmful with respect to economic consequences.

top20_propdmg_select <- top20_propdmg %>% 
  select(EVTYPE, propdmg) %>% 
  arrange(desc(propdmg))
agg_propdmg <- left_join(top20_propdmg_select, count_event, by="EVTYPE")
gt(agg_propdmg)

EVTYPE	propdmg	count
TORNADO	3212258.16	60652
FLASH FLOOD	1420124.59	54277
TSTM WIND	1335965.61	219940
FLOOD	899938.48	25326
THUNDERSTORM WIND	876844.17	82563
HAIL	688693.38	288661
LIGHTNING	603351.78	15754
THUNDERSTORM WINDS	446293.18	20843
HIGH WIND	324731.56	20212
WINTER STORM	132720.59	11433
HEAVY SNOW	122251.99	15708
WILDFIRE	84459.34	2761
ICE STORM	66000.67	2006
STRONG WIND	62993.81	3566
HIGH WINDS	55625.00	1533
HEAVY RAIN	50842.14	11723
TROPICAL STORM	48423.68	690
WILD/FOREST FIRE	39344.95	1457
FLASH FLOODING	28497.15	682
URBAN/SML STREAM FLD	26051.94	3392

Shoing the plot indicating which type are most harmful and frequent in terms of economic consequence. X axis indicate the property damage. Y axis indicates the type of events. The size of point indicates how frequent the event occurs.

agg_propdmg %>% 
  drop_na(propdmg) %>% 
  ggplot(aes(x=propdmg, y=EVTYPE, size=count, color=count)) +
  geom_point() +
  theme_bw()

Note that the echo = FALSE parameter was added to the code chunk to prevent printing of the R code that generated the plot.

Health and economic damage by severe weather events in US

TM

2023-01-28