Analysis of Avalanche Accidents and Fatalities in the United States from 1950 to 2022.
Data is available thanks to Colorado Avalanche Information Center (CAIC)
image credit: Matea Nikolina via Unsplash
I have been skiing for a very long time. As a child my father would put on ski movies after a long day on the slopes, most of the movies we would watch would have at least one scene of a professional skier or snowboarder riding down a slope as an avalanche is triggered. Much like quicksand I thought avalanches would be a very big worry as a grew older, but when one skis at a sanctioned ski resort avalanche dangers are slim to none if you are careful.
But I was still curious about them, I wanted to know when they actually occur and how they are triggered. I came across an expansive data set from the Colorado Avalanche Information Center and attempted to answer some of the questions I had.
The variables present in the data set include copious information about time, location, trigger type, travel mode of those stuck in the avalanche and fatalities.
I was most interested in the basic location, travel mode, fatalities and year.
a_year %>%
ggplot() +
geom_line(aes(x = YYYY, y = count, color = type), linewidth = 1) +
labs(title = "Have Avalanche Accidents and Deaths Increased Over Time?",
x = "Year",
y = "Total Events",
caption = "Data: Colorado Avalanche Information Center") +
scale_color_manual(values = c("#E69F00", "#0072B2"),name = "Type", labels = c("Death","Avalanche"))+
theme_minimal()
single = paste0(states2$full,"<br>",states2$count," Accident", sep="")
multiple = paste0(states2$full,"<br>",states2$count," Accidents", sep="")
mappedCounts <- states2 %>%
ggplot(aes(x, y,
group = group,
data_id = full)) +
geom_polygon_interactive(color = "grey45",
fill = "grey",
aes(tooltip = ifelse(count > 0, ifelse(count > 1, multiple, single), paste0(full,"<br>No Avalanche Accidents", sep="")))) +
labs(title="Avalanche Accidents",subtitle = "From April 1951 to May 2022")+
theme_minimal()+
theme(axis.title.x = element_blank(),
axis.title.y = element_blank(),
axis.text.x = element_blank(),
axis.ticks.x = element_blank(),
axis.text.y = element_blank(),
axis.ticks.y = element_blank())
graphedCount = states_count %>%
filter(count >0)%>%
ggplot()+
geom_col_interactive(aes(x = fct_reorder(full,count),
y = count,
tooltip = ifelse(count > 1, paste0(full,"<br>",count," Accidents", sep=""),paste0(full,"<br>",count," Accident", sep="")),
data_id = full))+
labs(x = "State", y = "Number of Accidents", caption = "Data: Colorado Avalanche Information Center")+
theme_minimal()+
coord_flip()
p = mappedCounts/graphedCount
girafe(ggobj = p, width_svg = 4.5, height_svg = 5.5,options = list(opts_sizing(rescale = TRUE, width = .75)))g <- list(
scope = 'usa',
projection = list(type = 'albers usa'),
showland = TRUE,
landcolor = toRGB("gray95"),
subunitcolor = toRGB("gray85"),
countrycolor = toRGB("gray85"),
countrywidth = 0.5,
subunitwidth = 0.5
)
fig <- plot_geo(recentAccidents, lat = ~lat, lon = ~lon)
fig <- fig %>% add_markers(
text = ~paste(Location,
paste("State:",state_full),
paste(Killed, PrimaryActivity, "Died"),
Date, sep = "<br />"),
color = ~factor(Killed), symbol = I("diamond"), size = I(8), hoverinfo = "text"
)
fig <- fig %>% layout(legend=list(title=list(text='<b> # of Deaths </b>')))
fig <- fig %>% layout(
title = 'Avalanche Deaths In United States <br /> (From January 2000 to May 2022)',
geo = g
)
fig#htmlwidgets::saveWidget(fig, "~/Documents/DS-Masters/CWD/Presentations/img/fig.html", selfcontained=TRUE, libdir = "lib")accidents %>%
filter(TravelMode %in% c("Foot", "Ski", "Snowboard", "Snowshoe", "Snowmobile")) %>%
group_by(TravelMode) %>%
summarize(total_accidents = n(), total_deaths = sum(Killed)) %>%
ggplot(aes(x = fct_reorder(TravelMode, desc(total_accidents)), y = total_accidents)) +
geom_col() +
labs(
title= "Which Mode of Transportation has the Highest Number of Accidents?",
x = "Travel Type",
y = "Number of Avalanches"
)+
theme_minimal()
travel_type %>%
ggplot() +
geom_col(aes(x = fct_reorder(TravelMode, desc(count)), y = count, fill = type), position = "dodge") +
labs(title = "Distribution of Accidents and Deaths for the Top Modes of Transportation?",
x = "Travel Mode",
y = "Number of Accidents") +
scale_fill_manual(values = c("darkgreen", "darkred"), name = "Type", labels = c("Avalanche", "Death"))+
theme_minimal()