In rafaelcb/NOAA: Explore NOAA earthquake data

This package is designed to facilitate the analysis of NOAA earthquake data from the Significant Earthquake Database. The package itself includes a sample of the data (eq_data), but its main functionality is in functions to clean up the raw data, which should be downloaded from the NOAA website, and visualize it in a convenient manner.

Cleaning up data

The data from the website is contained in textfiles that need to be tidied. For this, the functions eq_clean_data and eq_location_clean facilitate getting clean versions of the date and location of the earthquakes. The second function, in particular, cleans up the location name to remove redundant country data, while the first one adds a DATE column.

library(NOAA)
library(dplyr, warn.conflicts = FALSE)

eq_data %>% eq_clean_data() %>% eq_location_clean() %>% select(-YEAR, -MONTH, -DAY)

In particular, the resulting dataframe is useful for the visualization functions of the package.

Visualizing timelines of earthquakes

The package provides two geoms to easily visualize the timelines of earthquakes. The first one, geom_timeline generates a timeline of earthquakes for one or multiple countries, with aesthetic mappings that allow to control for size and color based on other variables.

library(ggplot2)
library(lubridate)

eq_data %>% eq_clean_data() %>% eq_location_clean() %>% 
    filter(lubridate::year(DATE) > 2000, COUNTRY %in% c("USA", "CHINA", "MEXICO")) %>% 
    ggplot(aes(x = DATE, y = COUNTRY, size = EQ_PRIMARY)) + geom_timeline() +
    ggtitle("Timeline of earthquakes in the USA, Mexico and China")

The geom_timeline_label function allows the user to additionally include labels for the most important events. To avoid cluttering, the number of labels is controlled using the n_max argument.

eq_data %>% eq_clean_data() %>% eq_location_clean() %>% 
    filter(lubridate::year(DATE) > 2000, COUNTRY %in% c("JAPAN", "CHILE", "MEXICO")) %>% 
    ggplot(aes(x = DATE, fill = TOTAL_DEATHS, size = EQ_PRIMARY)) + geom_timeline() + 
    geom_timeline_label(aes(label = LOCATION_NAME), n_max = 5) +
    ggtitle("Timeline of earthquakes in Japan, Chile and Mexico")

Additionally, a specific theme theme_eq_timeline object was added to improve the visualization of the data.

eq_data %>% eq_clean_data() %>% eq_location_clean() %>% 
    filter(lubridate::year(DATE) > 2000, COUNTRY %in% c("JAPAN", "CHILE", "MEXICO")) %>% 
    ggplot(aes(x = DATE, fill = TOTAL_DEATHS, size = EQ_PRIMARY, y = COUNTRY)) + geom_timeline() + 
    geom_timeline_label(aes(label = LOCATION_NAME), n_max = 5) + theme_eq_timeline +
    ggtitle("Timeline of earthquakes in Japan, Chile and Mexico")

Map genaration capabilities

Finally, the package has functions to easily map earthquakes. The function eq_map uses the leaflet package to create a beautiful visualization of the data. Additionally, a convenience function eq_create_label is included that allows to include information about the date, location and number of deaths as a result of the earthquake in popups that are triggered after clicking on a point in the map.

eq_data %>% eq_clean_data() %>% eq_location_clean() %>% 
    filter(lubridate::year(DATE) > 2000, COUNTRY == "CHINA") %>%
    mutate(pop_text = eq_create_label(.)) %>% 
    eq_map(annot_col = pop_text)