Plotting Geospatial Data: Vectors and Rasters

Author

Juliet Cohen

Published

November 25, 2025

Plotting Geospatial Data

In R, various geospatial packages exist for visualizing and manipulating spatial data:

  • sf
  • ggplot2
  • mapview
  • leaflet
  • tmap
  • terra

Visualizing vector data, raster data, or both overlaid can be accomplished in various ways. This resource provides example code and explanations for which packages are recommended to get the most out of your visualizations.

Which mapping package should I use?

Preference Package
Are you already familiar with ggplot for plotting, and you’re interested in a static visualization? ggplot2
Are you interested in highly tailorable map features and options to create either static or interactive visualizations? tmap
Are you interested in mapping raster data quickly with less code and less flexibility? terra::plot
Are you interested creating interactive maps quickly with high flexibility? leaflet
Are you interested in exploring spatial objects of any sort interactively? mapview

Vector Data: Points

sf, tmap and ggplot2 are great options for visualizing vector data, which is tabular data such as points, lines, and polygons.

Examples of vector file formats:

  • shapefile (.shp and auxillary files .shx, .dbf, .prj, etc.)
  • GeoPackage (.gpkg)
  • GeoParquet (.parquet)
  • GeoJSON (.geojson)

Load point data

Use the gbif API to download species observations for polar bears from the Global Biodiversity Information Facility.

Show the code 
# read in a list of items containing polar bear data, including metadata
pb_data <- occ_search(scientificName = "Ursus maritimus", 
                      limit = 300)

# subset the imported data to just the relevant dataframe and attributes
pb_obs <- pb_data$data %>% 
  select(decimalLongitude, 
         decimalLatitude, 
         year,
         country) %>%
  mutate(year = as.factor(year)) # year = categorical
  
# remove rows with NA in any col
pb_obs <- na.omit(pb_obs)

sf

Spatial data will not always already contain critical spatial metadata, so you may have to manually assign it using spatial operations before plotting. For example, point data may contain latitude and longitude coordinates into separate columns and may not come with a set coordinate reference system (CRS). sf can help create a geometry column from separate latitude and longitude columns and set the CRS to WGS84, EPSG:4326.

Show the code 
# convert separate longitude and latitude columns into 
# cohesive point geometries, and set the CRS
pb_spatial <- pb_obs %>% 
  sf::st_as_sf(coords = c("decimalLongitude", "decimalLatitude"),
               crs = st_crs(4326)) %>% 
              filter(st_is_valid(.))

Base R plot: points

Start with the basics: plot the point data using the native R function plot(), which does not include an interactive feature and is not highly tailorable.

Show the code 
plot(st_geometry(pb_spatial),
     main = "Polar Bear Observations",
     col = "black",
     pch = 16,
     axes = TRUE,
     xlab = "Longitude",
     ylab = "Latitude")

That static plot is pretty bland, and it has little context without a palette or a basemap.

Interactive Maps

Make an interactive map with the color of the points representing the year of the observation on a basemap. This can be done with either mapview or leaflet.

mapview

Show the code 
mapview(pb_spatial,
        zcol = "year",
        map.types = "Esri.NatGeoWorldMap",
        legend = TRUE,
        layer.name = "Polar Bear Observations")

The points are clickable when this is rendered locally, and a metadata window pops up for each observation.

leaflet

Show the code 
palette <- colorFactor(palette = 'viridis',
                       domain = pb_spatial$year)

leaflet(data = pb_spatial) %>%
  addProviderTiles("Esri.NatGeoWorldMap") %>% 
  addCircleMarkers(
    radius = 5,
    color = "black",  # point edges
    fillColor = ~palette(year),
    fillOpacity = 0.7,
    stroke = TRUE,
    weight = 1,  # point edge thickness
    popup = ~paste("Year:", year) # clickable points, show observation year
  ) %>%
  addLegend(
    "bottomright",
    pal = palette, 
    values = ~year,
    title = "Polar Bear Observations",
    opacity = 1
  )

ggplot: points on basemaps - static maps

sf and ggplot can be used in conjunction to plot the polar bear observations statically on a basemap. The default x and y gridlines are cohesive with latitude/longitude point data.

Show the code 
world <- rnaturalearth::ne_countries(scale = "medium", returnclass = "sf") 

ggplot(data = world) +
  geom_sf() +
  geom_sf(data = pb_spatial, 
          aes(fill = year),  # point color based on 'year'
          color = "black",  # point edges black
          shape = 21,
          size = 2, 
          alpha = 0.7) +  # transparency
  labs(title = "Polar Bear Observations",
       x = "Longitude",
       y = "Latitude",
       fill = "Year") +
  # limit map to polar bear habitat latitudes
  coord_sf(xlim = c(-180, 180), ylim = c(45, 90), expand = FALSE) +
  theme_minimal() + 
  theme(legend.position = "right")

ggplot can make more complex maps, too:

Show the code 
ggplot() +
  geom_sf(data = world, fill = "palegreen", color = "darkgreen") +  # Gray land with dark borders
  geom_sf(data = pb_spatial, 
          aes(fill = year),
          color = "black",
          shape = 21, 
          size = 2, 
          alpha = 0.7) +
  # limit the map to polar bear habitat latitudes & add more gridlines
  coord_sf(xlim = c(-180, 180),
           ylim = c(45, 90),
           expand = FALSE) +
  scale_x_continuous(breaks = seq(-180, 180, by = 10)) +
  scale_y_continuous(breaks = seq(45, 90, by = 10)) +
  labs(title = "Polar Bear Observations",
       subtitle = "CRS EPSG:4326",
       x = "Longitude",
       y = "Latitude",
       fill = "Year") +
  theme(panel.background = element_rect(fill = "lightblue"), # blue ocean
        plot.title = element_text(hjust = 0.5), # center the title
        plot.subtitle = element_text(hjust = 0.5),
        legend.position = "right",
        legend.box.background = element_rect(color = "black", 
                                             size = 0.5)) 
Warning: The `size` argument of `element_rect()` is deprecated as of ggplot2 3.4.0.
ℹ Please use the `linewidth` argument instead.

tmap: points on basemaps - static or interactive maps

tmap is specifically designed for mapping spatial data with many highly tailorable options, making it more customizable than ggplot. It recognizes spatial objects from sf, terra, and other geospatial packages. tmap can make both static and interactive maps, as it builds on ggplot and leaflet. More detailed basemaps, like those availble from leaflet and ESRI, are only an option in interactive mode for tmap. For large-scale static data, you can load in a simple world map to use as a basemap.

tmap allows for fine control over the locations of the title and legend. You can choose inside or outside the map, with values between 0-1 specified for the x and y position.

Make a static tmap:

Show the code 
# clarify the default mode is static plot
tmap_mode("plot")
ℹ tmap mode set to "plot".
Show the code 
data(World)

tm_shape(World) +
  tm_borders(col = "black", lwd = 0.5) +
  tm_fill(col = "white") +
  tm_shape(pb_spatial) +
  tm_dots(col = "year",
          palette = 'viridis',
          size = 0.1,
          border.col = "black",
          title = "Year") +
  tm_layout(
    bg.color = "lightblue",
    title = "Polar Bear\nObservations",
    frame = TRUE,
    title.position = c(0.01, 0.5),
    title.size = 1.2,
    legend.frame = TRUE,
    legend.position = c(0.01, 0.2)
  )

── tmap v3 code detected ───────────────────────────────────────────────────────
[v3->v4] `tm_tm_dots()`: migrate the argument(s) related to the scale of the
visual variable `fill` namely 'palette' (rename to 'values') to fill.scale =
tm_scale(<HERE>).[v3->v4] `tm_dots()`: use 'fill' for the fill color of polygons/symbols
(instead of 'col'), and 'col' for the outlines (instead of 'border.col').[tm_dots()] Argument `title` unknown.[v3->v4] `tm_layout()`: use `tm_title()` instead of `tm_layout(title = )`[tip] Consider a suitable map projection, e.g. by adding `+ tm_crs("auto")`.The visual variable "fill" of the layer "dotssymbols" contains a unique value. Therefore a categorical scale is applied (tm_scale_categorical).

Make an interactive tmap:

Show the code 
# set mode to interactive
tmap_mode("view")
ℹ tmap mode set to "view".
Show the code 
tm_shape(World) +
  tm_borders(col = "black", 
             lwd = 0.5) +
  tm_fill(col = "white", 
          alpha = 0.5) +
  tm_shape(pb_spatial) +
  tm_dots(col = "year",
          palette = "viridis",
          size = 0.1,
          border.col = "black",
          title = "Year") +
  tm_layout(bg.color = "lightblue",
            title = "Polar Bear Observations",
            title.size = 1.2,
            legend.frame = TRUE)

── tmap v3 code detected ───────────────────────────────────────────────────────
[v3->v4] `tm_polygons()`: use `fill_alpha` instead of `alpha`.[tm_dots()] Argument `title` unknown.[v3->v4] `tm_layout()`: use `tm_title()` instead of `tm_layout(title = )`The visual variable "fill" of the layer "dotssymbols" contains a unique value. Therefore a categorical scale is applied (tm_scale_categorical).

Vector Data: Polygons

Polygon data is composed of multiple points connected by lines to create closed shapes. Since there are many polar bears in Canada and Greenland, plot the polar bear observations on top of only Canada and Greenland polygons using both tmap and ggplot.

Show the code 
canada_greenland <- rnaturalearth::ne_countries(scale = "medium", 
                                                returnclass = "sf") %>% 
                                   filter(admin %in% c("Canada", "Greenland"))

We only want to plot the polar bear points that fit within these polygons, so execute a spatial join.

Show the code 
pb_canada_greenland <- st_join(pb_spatial, 
                               canada_greenland, 
                               join = st_within,
                               left = FALSE)

ggplot points and polygons

Show the code 
blue_palette <- RColorBrewer::brewer.pal(n = 2, name = "Blues")

ggplot(data = canada_greenland) +
  geom_sf(aes(fill = admin), 
          color = "black") +
  scale_fill_manual(values = blue_palette) +
  geom_sf(data = pb_canada_greenland,
          color = "red",
          size = 1) +
  theme_minimal() +
  labs(title = "Canada and Greenland Polar Bear Observations",
       fill = "Country",
       xlab = "Longitude",
       ylab = "Latitude")

tmap points and polygons

Show the code 
# static map setting, this is the default, but
# needs to be reset if previously set to "interactive"
tmap_mode("plot")
ℹ tmap mode set to "plot".
Show the code 
tm_shape(canada_greenland) + 
  tm_fill(col = "admin", 
          palette = blue_palette, 
          title = "Country") +
  tm_borders(col = "black") +
  tm_shape(pb_canada_greenland) +
  tm_dots(col = "red", size = 0.1) +
  tm_grid(lines = TRUE,
          col = "gray",
          labels.inside.frame = TRUE,
          n.x = 10, n.y = 10, # number of gridlines on x and y axes
          alpha = 0.5) +
  tm_layout(title = "Canada and Greenland\nPolar Bear Observations",
            title.size = 1,
            legend.title.size = 0.9,
            title.fontface = "bold",
            legend.text.size = 0.8,
            title.position = c(0.66, 0.31)) +
  tm_xlab("Longitude") +
  tm_ylab("Latitude")

── tmap v3 code detected ───────────────────────────────────────────────────────
[v3->v4] `tm_tm_polygons()`: migrate the argument(s) related to the scale of
the visual variable `fill` namely 'palette' (rename to 'values') to fill.scale
= tm_scale(<HERE>).[v3->v4] `tm_polygons()`: migrate the argument(s) related to the legend of the
visual variable `fill` namely 'title' to 'fill.legend = tm_legend(<HERE>)'[v3->v4] `tm_layout()`: use `tm_title()` instead of `tm_layout(title = )`

Base R plot: polygon

Add a polygon for polar bear habitat range from the International Union for Conservation of Nature (IUCN) Red List. This means we have 3 vector objects overlaid: polygons for country borders, points for polar bear observations, and 1 polygon for habitat.

Show the code 
# search for file anywhere within "course-materials" dir
hab_filename = "data_0.shp"
hab_fp <- list.files(path = here("course-materials"),
                 pattern = hab_filename,
                 recursive = TRUE,
                 full.names = TRUE)

pb_habitat = st_read(hab_fp)
# convert from a "simple feature collection" with 
# 15 fields to just the geometry
pb_habitat_poly <- st_geometry(pb_habitat)
Show the code 
plot(pb_habitat_poly,
     main = "Polar Bear Habitat Range",
     col = "lightyellow",
     axes = TRUE,
     xlab = "Latitude",
     ylab = "Longitude")

ggplot: points and polygons

Show the code 
ggplot(data = canada_greenland) +
  geom_sf(aes(fill = admin), 
          color = "black") +
  scale_fill_manual(values = blue_palette) +
  geom_sf(data = pb_canada_greenland,
          color = "red",
          size = 1) +
  geom_sf(data = pb_habitat_poly,
          fill = "yellow",
          color = "darkgoldenrod1",
          alpha = 0.2) +
  theme_minimal() +
  labs(title = "Canada and Greenland Polar Bear Observations",
       subtitle = "with habitat range",
       fill = "Country",
       xlab = "Longitude",
      ylab = "Latitude") +
  # limit map window: zoom into Canada and Greenland
  coord_sf(xlim = st_bbox(canada_greenland)[c("xmin", "xmax")],
           ylim = st_bbox(canada_greenland)[c("ymin", "ymax")])

tmap: points and polygons

Show the code 
tm_shape(canada_greenland) + 
  tm_fill(col = "admin", 
          palette = blue_palette, 
          title = "Country") +
  tm_borders(col = "black") +
  tm_shape(pb_canada_greenland) +
  tm_dots(col = "red", size = 0.1) +
  tm_grid(lines = TRUE,
          col = "gray",
          labels.inside.frame = TRUE,
          n.x = 10, n.y = 10,
          alpha = 0.5) +
  tm_shape(pb_habitat_poly) +
  tm_fill(col = "yellow", 
          alpha = 0.2, 
          title = "habitat") +
  tm_borders(col = "darkgoldenrod1") +
  tm_layout(title = "Canada and Greenland\nPolar Bear Observations\nwith Habitat Range",
            title.size = 1,
            title.fontface = "bold",
            legend.title.size = 0.9,
            legend.text.size = 0.8,
            title.position = c(0.66, 0.31)) +
  tm_xlab("Longitude") +
  tm_ylab("Latitude")
── tmap v3 code detected ───────────────────────────────────────────────────────
[v3->v4] `tm_tm_polygons()`: migrate the argument(s) related to the scale of
the visual variable `fill` namely 'palette' (rename to 'values') to fill.scale
= tm_scale(<HERE>).
[v3->v4] `tm_polygons()`: migrate the argument(s) related to the legend of the
visual variable `fill` namely 'title' to 'fill.legend = tm_legend(<HERE>)'
[v3->v4] `tm_polygons()`: use `fill_alpha` instead of `alpha`.
[v3->v4] `tm_polygons()`: migrate the argument(s) related to the legend of the
visual variable `fill` namely 'title' to 'fill.legend = tm_legend(<HERE>)'
[v3->v4] `tm_layout()`: use `tm_title()` instead of `tm_layout(title = )`

Raster Data

terra specializes in raster data processing, which are n-dimensional arrays.

Examples of file formats:

  • GeoTIFF (Tag Image File Format, .tif)
  • netCDF (Network Common Data Form, .nc)
  • PNG or JPEG images (.png, .jpg)

terra

Import a raster of sea ice for the Northern Hemisphere and plot it as simply as possible with terra::plot

Show the code 
ice_filename = "N_197812_concentration_v3.0.tif"
ice_fp <- list.files(path = here("course-materials"),
                 pattern = ice_filename,
                 recursive = TRUE,
                 full.names = TRUE)

arctic_ice <- terra::rast(ice_fp)

terra::plot(arctic_ice,
            main = "Arctic Sea Ice Concentration")

This data has a default palette; each cell is color-coded in shades of blue to white, where dark blue is 0% ice (open ocean) and white is 100% ice. You can view the default palette (“color table”) with terra::coltab

The CRS is projected and in units of meters, with each raster cell representing 25 km x 25 km. See CRS metadata here.

Make a histogram of the raster values using the base R hist to understand the numerical data distribution:

Show the code 
hist(arctic_ice,
     main = "Arctic Sea Ice Concentration Raster Values",
     xlab = "Values",
     ylab = "Frequency",
     col = "deepskyblue",
     border = "black")

In order to properly plot multiple spatial objects on top of one another, they must have the same CRS. Transform the CRS of the habitat polygon into the CRS of the raster for Arctic sea ice, then plot the habitat polygon onto the raster.

Show the code 
pb_habitat_arctic <- st_transform(pb_habitat_poly, st_crs(arctic_ice))

terra::plot(arctic_ice,
            main = "Sea Ice Concentration and Polar Bear Habitat")
terra::plot(pb_habitat_arctic, 
            add = TRUE, 
            border = "darkgoldenrod1", 
            col = adjustcolor("yellow", alpha.f = 0.2))

terra automatically defines the x and y axes ticks based on the spatial metadata of the raster.

ggplot: rasters and polygons

Scale the data values between 0-100 and convert the array into a dataframe, because ggplot only accepts tabular data. Assign a new palette using RColorBrewer that is similar to the color table associated with the terra plot above.

Show the code 
# reverse color palette to better match the default terra::plot palette values
blue_palette <- rev(RColorBrewer::brewer.pal(n = 9, name = "Blues"))

# scale the data 0-100:
# there are no NA values in this raster but na.rm = TRUE is good practice
range <- range(values(arctic_ice), na.rm = TRUE)
arctic_ice_scaled <- (arctic_ice-range[1]) / (range[2]-range[1]) * 100

arctic_ice_df <- terra::as.data.frame(arctic_ice_scaled,
                               cells = FALSE, # do not create index col
                               xy = TRUE)  %>% # include lat and long cols
                 dplyr::rename(ice_concentration = N_197812_concentration_v3.0)

ggplot() +
  geom_raster(data = arctic_ice_df,
              aes(x = x, y = y,
                  fill = ice_concentration)) +
  scale_fill_gradientn(colors = blue_palette) +
  geom_sf(data = pb_habitat_arctic,
          fill = "yellow",
          color = "darkgoldenrod1",
          size = 0.2,
          alpha = 0.1) +
  coord_sf(default_crs = st_crs(pb_habitat_arctic)) +
  theme(panel.grid = element_blank(),
        axis.text = element_blank(),
        axis.ticks = element_blank(),
        panel.background = element_rect(fill = "grey", color = NA),
        plot.background = element_rect(fill = "grey", color = NA)) +
  labs(title = "Sea Ice Concentration\nand Polar Bear Habitat Range",
       subtitle = "Proj CRS: NSIDC Sea Ice Polar Stereographic North",
       fill = "Sea Ice\nConcentration",
       x = element_blank(),
       y = element_blank())

Note that ggplot does not automatically derive the units of meters from the projected CRS from the spatial metadata. Instead, it uses 4326 by default. As a result, we mask the axes ticks. Axes ticks can manually be defined with scale_x_continuous and scale_y_continuous

tmap: rasters and polygons

Show the code 
# tmap_mode("plot")
# 
# #tmap_options(max.categories = 9)
# 
# tm_shape(arctic_ice_scaled) +
#   tm_raster(palette = blue_palette,
#             title = "Sea Ice\nConcentration",
#             style = "cont",
#             breaks = seq(0, 100, length.out = 11),
#             midpoint = NA) +
# tm_shape(pb_habitat_arctic) +
#   tm_polygons(col = "yellow",
#               border.col = "darkgoldenrod1",
#               lwd = 1,
#               alpha = 0.1) +
# tm_graticules(n.x = 5, n.y = 5, 
#               labels.show = TRUE, 
#               labels.size = 0.6,
#               alpha = 0.3) +
# tm_layout(title = "Sea Ice\nConcentration\nand Polar Bear\nHabitat Range",
#           main.title.size = 0.8,
#           title.fontface = "bold",
#           legend.outside = TRUE,
#           legend.title.size = 1,
#           legend.outside.position = "right",
#           inner.margins = c(0.1, 0.1, 0.1, 0.1)) +
# tm_scale_bar(position = c("left", "bottom"))

Data Citations

Dataset Citation
GBIF, polar bear observation points GBIF.org (01 September 2024) GBIF Occurrence Download https://doi.org/10.15468/dl.79778w
IUCN Red List, polar bear range polygon IUCN. 2024. The IUCN Red List of Threatened Species. Version 2024-1. https://www.iucnredlist.org. Accessed on Septmeber 5, 2024.
NSIDC, sea ice concentration raster Fetterer, F., Knowles, K., Meier, W. N., Savoie, M. & Windnagel, A. K. (2017). Sea Ice Index. (G02135, Version 3). [Data Set]. Boulder, Colorado USA. National Snow and Ice Data Center. https://doi.org/10.7265/N5K072F8. [describe subset used if applicable]. Date Accessed 09-13-2024.

NSIDC sea ice concentration metadata