Exploring the package {geodata}

Exploring the various datasets available publicly and freely to use and analyze with R, {terra} and {sf} with the {geodata} package

Maps

Geocomputation

{geodata}

Author

Aditya Dahiya

Published

January 27, 2025

Exploring the {geodata} package

The {geodata} package (Hijmans et al. 2024) with {sf} (Pebesma and Bivand 2023) and {terra} (Hijmans 2024)

Code

# Data Import and Wrangling Tools
library(tidyverse)            # All things tidy
library(sf)                   # Handling simple features in R
library(terra)                # Handling rasters in R
library(tidyterra)            # Rasters with ggplot2

# Final plot tools
library(scales)               # Nice Scales for ggplot2
library(fontawesome)          # Icons display in ggplot2
library(ggtext)               # Markdown text in ggplot2
library(showtext)             # Display fonts in ggplot2
library(colorspace)           # Lighten and Darken colours

# Package to explore
library(geodata)              # Geospatial Data

bts = 12 # Base Text Size
sysfonts::font_add_google("Saira Condensed", "body_font")
showtext::showtext_auto()
theme_set(
  theme_minimal(
    base_size = bts,
    base_family = "body_font"
  ) +
    theme(
      text = element_text(
        colour = "grey20",
        lineheight = 0.3,
        margin = margin(0,0,0,0, "pt")
      )
    )
)

# Some basic caption stuff
# A base Colour
bg_col <- "white"
seecolor::print_color(bg_col)

# Colour for highlighted text
text_hil <- "grey30"
seecolor::print_color(text_hil)

# Colour for the text
text_col <- "grey20"
seecolor::print_color(text_col)


# Caption stuff for the plot
sysfonts::font_add(
  family = "Font Awesome 6 Brands",
  regular = here::here("docs", "Font Awesome 6 Brands-Regular-400.otf")
)
github <- "&#xf09b"
github_username <- "aditya-dahiya"
xtwitter <- "&#xe61b"
xtwitter_username <- "@adityadahiyaias"
social_caption_1 <- glue::glue("<span style='font-family:\"Font Awesome 6 Brands\";'>{github};</span> <span style='color: {text_hil}'>{github_username}  </span>")
social_caption_2 <- glue::glue("<span style='font-family:\"Font Awesome 6 Brands\";'>{xtwitter};</span> <span style='color: {text_hil}'>{xtwitter_username}</span>")
plot_caption <- paste0(
  "**Data**: {geodata} package  ",
  "  |  **Code:** ", 
  social_caption_1, 
  " |  **Graphics:** ", 
  social_caption_2
  )
rm(github, github_username, xtwitter, 
   xtwitter_username, social_caption_1, 
   social_caption_2)

Exploring the average monthly temperature climate data for Greenland

This code visualizes monthly average temperature data for Greenland using geospatial data and ggplot2 in R, leveraging key functions from the {geodata} package. The geodata::gadm() function is used to download Greenland’s administrative boundaries as vector data, which is processed into a simple feature object with sf::st_as_sf(). The geodata::worldclim_country() function retrieves average temperature climate data for Greenland. The climate data is aggregated for lower resolution using terra::aggregate(), cropped to Greenland’s boundaries with terra::crop(), and masked with terra::mask() to match the vector boundary. It is then reprojected to a North Pole Stereographic projection using terra::project(). A faceted ggplot ( Figure 1 ) is created to visualize temperature data across all 12 months, with color gradients representing temperature values, styled with a minimal theme.

Code

# Get Greenland's boundaries
greenland_vector <- geodata::gadm(
  country = "Greenland",
  path = tempdir(),
  resolution = 2
) |> 
  st_as_sf()

vec1 <- greenland_vector |> 
  st_union() |> 
  st_as_sf()

# ggplot(greenland_vector) +
#   geom_sf()

# Get Average Temperature Climate Data for Greenland
df1 <- worldclim_country(
  country = "Greenland",
  var = "tavg",
  path = tempdir()
)

# Studying the layers: there are 12 - for the 12 months
df1

df1 |> crs() |> str_view()

# Lower resolution for initial plots, and crop by Vector Map
df2 <- df1 |> 
  terra::aggregate(fact = 10, fun = mean) |>
  terra::crop(vec1) |> 
  terra::mask(vec1)

# Project into CRS: North Pole Stereographic
df3 <- df2 |> 
  project("EPSG:3413", method = "bilinear")

strip_labels <- month.name
names(strip_labels) <- names(df2)

g <- ggplot() +
  geom_spatraster(
    data = df3
  ) +
  scale_fill_viridis_c(
    na.value = "transparent",
    labels = function(x) paste0(x, "°C"),
  ) +
  facet_wrap(
    ~lyr,
    labeller = labeller(lyr = strip_labels),
    ncol = 3,
    nrow = 4
  ) +
  coord_sf(
    clip = "off"
  ) +
  labs(
    title = "Monthly Average Temperature in Greenland",
    fill = "Average\ntemperature",
    caption = plot_caption
  ) +
  theme_minimal(
    base_family = "body_font",
    base_size = bts * 3
  ) +
  theme(
    axis.text = element_text(
      size = bts, 
      margin = margin(0,0,0,0, "pt")
    ),
    axis.ticks.length = unit(0, "pt"),
    strip.text = element_text(
      margin = margin(0,0,0,0, "pt")
    ),
    panel.grid = element_line(
      linewidth = 0.1,
      linetype = 3,
      colour = "grey80"
    ),
    legend.position = "right",
    panel.spacing = unit(5, "pt"),
    panel.background = element_rect(
      fill = "transparent",
      colour = "transparent"
    ),
    
    # Legend Corrections
    legend.title = element_text(
      margin = margin(0,0,3,0, "pt"),
      lineheight = 0.3,
      hjust = 0.5
    ),
    legend.text = element_text(
      margin = margin(0,0,0,2, "pt"),
      size = 18
    ),
    legend.margin = margin(0,0,0,0, "pt"),
    legend.box.margin = margin(0,0,0,0, "pt"),
    legend.key.width = unit(4, "pt"),
    legend.key.height = unit(30, "pt"),
    plot.caption = element_textbox(
      hjust = 0.5,
      size = bts * 2
    ),
    plot.title = element_text(
      size = bts * 4,
      hjust = 0.5
    ),
    plot.title.position = "plot"
  )
  

ggsave(
  plot = g,
  filename = here::here(
    "geocomputation", "images", "geodata_package_1.png"
  ),
  height = 1800,
  width = 2000,
  units = "px",
  bg = "white"
)

Figure 1: Monthly Average Temperatures Across Greenland: showcasing temperature gradients for each month using a North Pole Stereographic projection.

Getting Elevation data with {geodata} and plotting sea level rise

This code demonstrates the simulation of various sea level rise scenarios using elevation data and visualizes the resulting impact on global landmass through raster manipulation and plotting. It leverages the elevation_global() function to obtain global elevation raster data, with a resolution of 10 arc-seconds. The if_else() function is used to simulate sea level rise scenarios by adjusting elevation data based on thresholds (1m, 10m, 50m, 100m, and 200m). The manipulated elevation data is then stored as layers in a raster object created with rast() from the terra package.

The raster is reprojected into an aesthetically pleasing projection (ESRI:54030) using project() for better visualization. The final map is plotted with ggplot(), incorporating the geom_spatraster() layer to display raster data and facet_wrap() to present the different scenarios.

Code

# Getting blobale elevation raster data
ge_raw <- elevation_global(
  res = 10,
  path = tempdir()
)

# Simulating Sea Leave Rise by different heights
elev_tibble <- tibble(
  base = values(ge_raw),
  rise_1 = if_else(base < 1, NA, base),
  rise_10 = if_else(base < 10, NA, base),
  rise_40 = if_else(base < 50, NA, base),
  rise_80 = if_else(base < 100, NA, base),
  rise_100 = if_else(base < 200, NA, base)
)

# Make a blank raster of World Map size
ge1 <- rast(
  nrow = dim(ge_raw)[1],
  ncol = dim(ge_raw)[2],
  nlyrs = ncol(elev_tibble)
)
# Add values from the tibble to different layers of 
# the newly created raster
values(ge1) <- as.matrix(elev_tibble)

# Reproject Raster into a nice projection 
ge1 <- ge1 |> 
  project("ESRI:54030")

# Earlier Attempts that did not work
# ge1 <- ge_raw
# sea_level_rise = 1       # In metres
# values(ge1) <- if_else(
#   values(ge_raw) < sea_level_rise,
#   NA,
#   values(ge_raw)
#   )
# 
# paste0("ge", sea_level_rise) |> 
#   assign(ge_raw)
# 

strip_labels <- c(
  "Current Sea Level",
  paste0("Rise by ", 
         c(1, 10, 50, 100, 200), 
         " metres")
)
names(strip_labels) <- paste0("lyr.", 1:6)
  
g <- ggplot() +
  geom_spatraster(
    data = ge1
  ) +
  facet_wrap(
    ~lyr, 
    ncol = 2,
    labeller = labeller(lyr = strip_labels)
    ) +
  scale_fill_wiki_c(
    na.value = "lightblue"
  ) +
  coord_sf(clip = "off") +
  labs(
    title = "World Map: scenarios with rise in Sea Levels",
    subtitle = "Simulating different levels of sea level rise, and effect of landmass using simple raster arithmetic",
    caption = plot_caption,
    fill = "Elevation above sea level (m)"
  ) +
  theme_minimal(
    base_size = 2 * bts,
    base_family = "body_font"
  ) +
  theme(
    plot.margin = margin(0,0,0,0, "pt"),
    panel.spacing = unit(2, "pt"),
    legend.position = "bottom",
    legend.title.position = "top",
    panel.grid = element_blank(),
    plot.caption = element_textbox(
      hjust = 0.5,
      margin = margin(5,0,5,0, "pt"),
      size = bts
    ),
    plot.title = element_text(
      margin = margin(15,0,0,0, "pt"),
      hjust = 0.5,
      face = "bold"
    ),
    plot.subtitle = element_text(
      margin = margin(5,0,0,0, "pt"),
      hjust = 0.5,
      size = bts * 1.5
    ),
    strip.text = element_text(
      margin = margin(2,0,-1,0, "pt"),
      face = "bold"
    ),
    legend.key.height = unit(5, "pt"),
    legend.box.margin = margin(-10,0,3,0, "pt"),
    legend.margin = margin(-10,0,0,0, "pt"),
    legend.title = element_text(
      margin = margin(0,0,2,0, "pt"),
      hjust = 0.5
    ),
    legend.text = element_text(
      margin = margin(1,0,0,0, "pt"),
      hjust = 0.5,
      size = bts
    )
  )

ggsave(
  plot = g,
  filename = here::here(
    "geocomputation", "images", "geodata_package_2.png"
  ),
  height = 1150,
  width = 1000,
  units = "px",
  bg = "lightblue"
)

An elevation map of the world plotted using geom_spatraster() which shows 6 facets - each with a different scenario of sea level rise with global warming - current levels, 1 m, 10 m, 50 m, 100 m and 200 m - drawn with simple raster arithmetic. The world is projected in Robinson Projection (ESRI:54030) to make is aesthetically pleasing.

Doing the same map for Europe

This code extends the sea level rise simulation to the European region by refining the raster data and restricting it to the continent’s boundaries. It first retrieves global elevation data using elevation_global(), then defines a bounding box to exclude distant islands and ensure a focused analysis on mainland Europe.

To obtain vector data for European countries, ne_countries() from the rnaturalearth package is used, selecting only relevant attributes. Unlike the previous global analysis, this visualization applies the ETRS89 / LAEA Europe projection (EPSG:32633) using project() to maintain spatial accuracy for the European context. The final visualization integrates raster data using geom_spatraster() and overlays country borders (shown in white colour) with geom_sf(), providing a detailed view of how landmass is affected by rising sea levels.

Code

# Getting global elevation raster data
ge_raw <- elevation_global(
  res = 10,
  path = tempdir()
)

# Create a bounding box to remove far off islands in Europe
bounding_box <- st_bbox(
  c(xmin = -25, xmax = 45, ymin = 31, ymax = 70),
  crs = "EPSG:4326"
)

# Get Vector Data on European Countries
europe <- rnaturalearth::ne_countries(
  continent = "Europe",
  returnclass = "sf",
  scale = "medium"
) |> 
  select(name, iso_a3, geometry) |> 
  filter(!(name %in% c("Russia"))) |> 
  st_crop(bounding_box)

ggplot(europe) +
  geom_sf()

whole_europe <- europe |>
  st_union() |> 
  st_as_sf()

# Check the map
ggplot(whole_europe) +
  geom_sf()

# Crop the global Elevation raster
ge_europe <- ge_raw |> 
  terra::crop(whole_europe) |> 
  terra::mask(whole_europe)

ggplot() +
  geom_spatraster(data = ge_europe)

# Simulating Sea Leave Rise by different heights
elev_tibble <- tibble(
  base = values(ge_europe),
  rise_1 = if_else(base < 1, NA, base),
  rise_10 = if_else(base < 10, NA, base),
  rise_40 = if_else(base < 50, NA, base),
  rise_80 = if_else(base < 100, NA, base),
  rise_100 = if_else(base < 200, NA, base)
)

# Make a blank raster of World Map size
ge1 <- rast(
  nrow = dim(ge_europe)[1],
  ncol = dim(ge_europe)[2],
  nlyrs = ncol(elev_tibble),
  extent = ext(ge_europe),
  resolution = res(ge_europe)
)
# Add values from the tibble to different layers of 
# the newly created raster
values(ge1) <- as.matrix(elev_tibble)

# Test the ge1
# ge1
# ge_europe

# Reproject Raster into ETRS89 / LAEA Europe Projection for Europe 
ge_projected <- ge1 |> 
  project("EPSG:32633")

# Earlier Attempts that did not work
# ge1 <- ge_raw
# sea_level_rise = 1       # In metres
# values(ge1) <- if_else(
#   values(ge_raw) < sea_level_rise,
#   NA,
#   values(ge_raw)
#   )
# 
# paste0("ge", sea_level_rise) |> 
#   assign(ge_raw)
# 

strip_labels <- c(
  "Current Sea Level",
  paste0("Rise by ", 
         c(1, 10, 50, 100, 200), 
         " metres")
)
names(strip_labels) <- paste0("lyr.", 1:6)
  
g <- ggplot() +
  geom_spatraster(
    data = ge_projected
  ) +
  geom_sf(
    data = europe,
    colour = "white",
    fill = "transparent",
    linewidth = 0.1
  ) +
  facet_wrap(
    ~lyr, 
    ncol = 2,
    labeller = labeller(lyr = strip_labels)
    ) +
  scale_fill_wiki_c(
    na.value = "lightblue"
  ) +
  coord_sf(clip = "off") +
  labs(
    title = "Europe Map with rising in Sea Levels",
    subtitle = "Simulating different levels of sea level rise, and effect of landmass using simple raster arithmetic",
    caption = plot_caption,
    fill = "Elevation above sea level (m)"
  ) +
  theme_minimal(
    base_size = 2 * bts,
    base_family = "body_font"
  ) +
  theme(
    plot.margin = margin(0,0,0,0, "pt"),
    panel.spacing = unit(2, "pt"),
    legend.position = "bottom",
    legend.title.position = "top",
    panel.grid = element_blank(),
    plot.caption = element_textbox(
      hjust = 0.5,
      margin = margin(5,0,5,0, "pt"),
      size = bts
    ),
    plot.title = element_text(
      margin = margin(15,0,0,0, "pt"),
      hjust = 0.5,
      face = "bold"
    ),
    plot.subtitle = element_text(
      margin = margin(3,0,1,0, "pt"),
      hjust = 0.5,
      size = bts * 1.5
    ),
    strip.text = element_text(
      margin = margin(2,0,-9,0, "pt"),
      face = "bold"
    ),
    legend.key.height = unit(5, "pt"),
    legend.box.margin = margin(-10,0,3,0, "pt"),
    legend.margin = margin(-10,0,0,0, "pt"),
    legend.title = element_text(
      margin = margin(0,0,2,0, "pt"),
      hjust = 0.5
    ),
    legend.text = element_text(
      margin = margin(1,0,0,0, "pt"),
      hjust = 0.5,
      size = bts
    ),
    axis.text = element_blank(),
    axis.ticks = element_blank(),
    axis.ticks.length = unit(0, "pt")
  )

ggsave(
  plot = g,
  filename = here::here(
    "geocomputation", "images", "geodata_package_3.png"
  ),
  height = 1150,
  width = 1000,
  units = "px",
  bg = "lightblue"
)

Simulating Sea Level Rise in Europe: This map visualizes the impact of rising sea levels on European landmass using raster-based elevation data. Different scenarios, ranging from current sea levels to rises of 1m, 10m, 50m, 100m, and 200m, are represented to illustrate potential changes in geography. The analysis is tailored to the European context with an accurate ETRS89 / LAEA projection. The existing borders of nations are shown in white.

Exploring Crop data from {geodata}

This R script uses open-source land cover data from the {geodata} package to analyze and visualize cropland distribution across Haryana’s subdistricts (tehsils). The dataset, derived from the ESA WorldCover dataset at a 0.3-second resolution, represents the fraction of cropland in each cell. The analysis starts by downloading global cropland data using geodata::cropland(). The Haryana subdistrict boundary shapefile is read using {sf} via read_sf(), cleaned with {janitor}, and transformed to EPSG:4326. The cropland raster is cropped and masked to Haryana’s boundaries using {terra} functions crop() and mask(). Raster extraction (terra::extract()) is performed to compute the average cropland percentage per subdistrict. The results are merged into the spatial data (left_join()) and ranked. Visualization is done using {ggplot2}, where an inset bar chart ranks subdistricts by cropland percentage.

Code

globalcrop <- geodata::cropland(
  source = "worldcover",
  path = tempdir()
)

haryana_vec <- read_sf(
  here::here(
    "data", "haryana_map",
    "HARYANA_SUBDISTRICT_BDY.shp"
  )
) |> 
  janitor::clean_names() |> 
  mutate(
    across(
      .cols = c(district, tehsil),
      .fns = function(x){
        str_replace_all(x, "\\|", "I") |> 
          str_replace_all(">", "A") |> 
          str_replace_all("@", "U") |> 
          str_to_title()
      }
    )
  ) |> 
  st_transform("EPSG:4326") |> 
  mutate(id = row_number()) |> 
  relocate(id)

haryana_vec_boundary <- st_union(haryana_vec) |> 
  st_as_sf()

haryana_crop <- globalcrop |> 
  terra::crop(haryana_vec) |> 
  terra::mask(haryana_vec)

# Computing percentage area of each subdistrict that is Cropland
df1 <- haryana_crop |> 
  terra::extract(haryana_vec) |> 
  as_tibble() |> 
  group_by(ID) |> 
  summarise(crop_perc = mean(cropland)) |> 
  arrange(desc(crop_perc)) |> 
  janitor::clean_names() |> 
  left_join(haryana_vec |> st_drop_geometry() |> 
              select(id, district, tehsil)) |> 
  mutate(
    rank = row_number(),
    facet_var = if_else(rank <= 39, "1", "2")
  )

# Add Ranks to display
haryana_vec <- haryana_vec |> 
  left_join(df1)


g_inset <- df1 |>
  mutate(
    rank = fct(
      as.character(rank),
      levels = paste0(77:1)
    )
  ) |> 
  ggplot(
    mapping = aes(
      x = crop_perc, 
      y = rank
    )
  ) +
  geom_col(
    width = 0.6,
    fill = alpha("grey20", 0.5)
  ) +
  geom_text(
    mapping = aes(
      label = paste0(
        tehsil
      ),
      x = 0
    ),
    nudge_x = -0.05,
    hjust = 1,
    size = 4,
    lineheight = 0.3,
    family = "body_font",
    colour = "grey20"
  ) +
  geom_label(
    mapping = aes(label = rank, x = 0.02),
    family = "body_font",
    size = 4,
    label.size = unit(0.1, "pt"),
    label.padding = unit(0.1, "lines")
  ) +
  geom_text(
    mapping = aes(label = paste0(
      round(100*crop_perc, 1), "%")),
    family = "body_font",
    size = 4,
    hjust = 1,
    nudge_x = -0.03,
    colour = "white"
  ) +
  coord_cartesian(clip = "off") +
  facet_wrap(~facet_var, nrow = 1, scales = "free_y") +
  labs(
    x = NULL, y = NULL
  ) +
  scale_x_continuous(
    labels = label_percent(),
    expand = expansion(c(0.3, 0)),
    breaks = c(0, seq(0.1, 0.9, 0.2))
  ) +
  theme(
    axis.text.y = element_blank(),
    strip.text = element_blank(),
    panel.grid = element_blank()
  )

Code

g_base <- ggplot() +
  geom_spatraster(
    data = haryana_crop
  ) +
  paletteer::scale_fill_paletteer_c(
    "grDevices::Blue-Yellow",
    direction = -1,
    na.value = "transparent",
    trans = "exp"
  ) +
  geom_sf(
    data = haryana_vec,
    colour = "white",
    fill = "transparent"
  ) +
  geom_sf(
    data = haryana_vec_boundary,
    colour = "grey10",
    linewidth = 0.5,
    fill = "transparent"
  ) +
  geom_sf_label(
    data = haryana_vec,
    mapping = aes(label = rank),
    size = 2,
    colour = "grey10",
    fill = alpha("white", 0.9),
    label.size = NA,
    label.r = unit(0.1, "lines"),
    label.padding = unit(0.1, "lines")
  ) +
  coord_sf(clip = "off") +
  labs(
    title = "Cropland in different Sub-districts of Haryana (India)",
    subtitle = "Using data from ESA WorldCover, showing percentage area of each subdivision that is cropland.",
    caption = plot_caption,
    fill = "Percentage area in each pixel that is cropland",
    x = NULL, y = NULL
  ) +
  theme_minimal(
    base_size = bts * 1.5,
    base_family = "body_font"
  ) +
  theme(
    text = element_text(
      colour = "grey20",
      lineheight = 0.3
    ),
    plot.margin = margin(0,0,0,0, "pt"),
    legend.position = "inside",
    legend.position.inside = c(1, 0),
    legend.direction = "horizontal",
    legend.title.position = "top",
    legend.justification = c(1, 1),
    panel.grid = element_line(
      colour = "grey50",
      linetype = 3,
      linewidth = 0.2
    ),
    plot.caption = element_textbox(
      hjust = 0.5,
      margin = margin(35,0,5,0, "pt"),
      size = bts
    ),
    plot.title.position = "plot",
    plot.title = element_text(
      margin = margin(10,0,0,0, "pt"),
      hjust = 0.5,
      face = "bold",
      size = bts * 3
    ),
    plot.subtitle = element_text(
      margin = margin(3,0,1,0, "pt"),
      hjust = 0.5,
      size = bts * 2
    ),
    legend.key.height = unit(5, "pt"),
    legend.key.width = unit(35, "pt"),
    legend.box.margin = margin(10,5,3,0, "pt"),
    legend.margin = margin(10,5,0,0, "pt"),
    legend.title = element_text(
      margin = margin(0,0,2,0, "pt"),
      hjust = 0.5
    ),
    legend.text = element_text(
      margin = margin(1,0,0,0, "pt"),
      hjust = 0.5
    ),
    axis.text.y = element_text(
      margin = margin(0,0,0,180, "pt")
    ),
    axis.ticks = element_blank(),
    axis.ticks.length = unit(0, "pt")
  )

library(patchwork)
g <- g_base +
  inset_element(
    p = g_inset,
    left = -0.1, right = 0.47,
    bottom = 0, top = 0.91,
    align_to = "full"
  )

ggsave(
  plot = g,
  filename = here::here(
    "geocomputation", "images", 
    "geodata_package_4.png"
  ),
  height = 1300,
  width = 1800,
  units = "px",
  bg = "white"
)

This map visualizes the percentage of cropland across Haryana’s tehsils using high-resolution ESA WorldCover data. Each subdistrict is ranked by cropland coverage, with darker shades indicating higher agricultural presence. An inset chart provides a ranked comparison of tehsils, offering a clear insight into regional variations in cropland distribution. Data sourced from ESA WorldCover (CC BY 4.0).

References

Hijmans, Robert J. 2024. “Terra: Spatial Data Analysis.” https://CRAN.R-project.org/package=terra.

Hijmans, Robert J., Márcia Barbosa, Aniruddha Ghosh, and Alex Mandel. 2024. “Geodata: Download Geographic Data.” https://CRAN.R-project.org/package=geodata.

Pebesma, Edzer, and Roger Bivand. 2023. “Spatial Data Science: With Applications in r.” https://doi.org/10.1201/9780429459016.