saudimap 🇸🇦

I couldn’t really find a good source of map data for Saudi Arabia’s administrative boundaries, so I created my own from OpenStreetMap data using the {osmdata} package. In the data/ folder, you will find:

1. Dataset containing the regions and governorates of Saudi Arabia as sf objects, along with population data where available. Load the dataset in R using

   library(sf)
   #> Linking to GEOS 3.13.0, GDAL 3.8.5, PROJ 9.5.1; sf_use_s2() is TRUE
   load("data/ksamaps.Rdata")

2. GeoJSON files for the same, if you prefer these instead.

Features

There are two sf objects:

1. reg_sf: the 13 regions of Saudi Arabia
2. gov_sf: the 149 governorates of Saudi Arabia

Data

# What's in our data set?
tibble::glimpse(gov_sf)
#> Rows: 149
#> Columns: 7
#> $ governorate       <chr> "Dammam", "Khobar", "Abqaiq", "Al Ahsa", "Al Udayd",…
#> $ population        <chr> "903597", "578500", "53444", "1063112", NA, "52340",…
#> $ population.date   <chr> "2010", "2010", "2010", "2010", NA, "2010", "2022", …
#> $ source            <chr> "covid19 saudi MOH open data", "covid19 saudi MOH op…
#> $ source.population <chr> "stats.gov.sa", "stats.gov.sa", "stats.gov.sa", "sta…
#> $ region            <chr> "Eastern", "Eastern", "Eastern", "Eastern", "Eastern…
#> $ geometry          <MULTIPOLYGON [°]> MULTIPOLYGON (((50.07442 26..., MULTIPO…

Plot

library(ggplot2)
ggplot() +
  geom_sf(data = gov_sf, aes(fill = region), alpha = 0.65) +
  geom_sf(data = reg_sf, fill = NA, linewidth = 0.5, col = "black") +
  ggrepel::geom_label_repel(
    data = reg_sf,
    aes(label = region, geometry = geometry),
    size = 2.7,
    stat = "sf_coordinates",
    max.overlaps = Inf,
    min.segment.length = 0,
    segment.size = 0.3,
    segment.curvature = 0.1,
    force = 5
  ) +
  scale_fill_viridis_d(option = "mako", guide = "none") +
  theme_void()

Creating the datasets

The code used to create the datasets is in R/create_maps.R. The main steps were:

1. Use {osmdata} to pull all admin boundaries for Saudi Arabia from OpenStreetMap, filtering for admin levels 4 (regions) and 5/6 (governorates).

2. Intersect the polygons with the overall boundary of Saudi Arabia to remove any spillover from neighboring countries.

3. Clean up the data by removing non-polygon geometries and ensuring all geometries are valid.

4. Wrangle the data into two separate datasets: one for regions and one for governorates, including relevant attributes like name and population.

5. The governorate dataset needed to be matched with the correct region. This was the difficult bit, since the names are not exact. For e.g. “Khubar” and “Khobar” refer to the same place, and so to “Medina” and “Al-Madina” and “Al Medina”. This needed to be done manually, and for this I created the data/governorates.csv file. If there’s a standardised version of names and regions that would have been helpful, but I don’t know where to find that.

Source

As mentioned, the data comes from OpenStreetMap, which is crowd-sourced and open data. According to the Wiki, the KSA data was obtained from official government sources, namely the Saudi MOH Open Data (seems related to Covid-19 times).