Accessing Spatial and Population Data from Statistics Finland OGC api

Introduction

The geofi package provides tools to access spatial data from Statistics Finland’s OGC API, including administrative boundaries, population data by administrative units, and population data by statistical grid cells. This vignette demonstrates how to use the package’s core functions to:

Retrieve Finnish administrative area polygons (e.g., municipalities, regions).
Fetch population data linked to administrative units.
Access population data for statistical grid cells.

Unlike some other spatial data APIs, no API key is required to access Statistics Finland’s OGC API, making it straightforward to get started. The package handles pagination, spatial filtering, and coordinate reference system (CRS) transformations, delivering data as sf objects compatible with the sf package for spatial analysis and visualization.

Package Overview

The geofi package includes the following key functions for accessing Statistics Finland data:

ogc_get_statfi_area(): Retrieves administrative area polygons (e.g., municipalities, wellbeing areas) for specified years, scales, and tessellation types.
ogc_get_statfi_area_pop(): Fetches administrative area polygons with associated population data, pivoted into a wide format.
ogc_get_statfi_statistical_grid(): Retrieves population data for statistical grid cells at different resolutions (1km or 5km).
fetch_ogc_api_statfi(): An internal function that handles low-level API requests and pagination (not typically called directly by users).

All functions return spatial data as sf objects, making it easy to integrate with spatial analysis workflows in R.

Step 1: Retrieving Administrative Area Polygons

The ogc_get_statfi_area() function retrieves polygons for Finnish administrative units, such as municipalities (kunta), wellbeing areas (hyvinvointialue), or regions (maakunta). You can customize the output with parameters like:

year: The year of the boundaries (2020–2022).
scale: Map resolution (1:1,000,000 or 1:4,500,000).
tessellation: Type of administrative unit (e.g., kunta, hyvinvointialue).
crs: Coordinate reference system (EPSG:3067 or EPSG:4326).
limit: Maximum number of features (or NULL for all).
bbox: Bounding box for spatial filtering.

Example: Downloading Municipalities

Fetch all municipalities for 2022 at the 1:4,500,000 scale:

munis <- ogc_get_statfi_area(year = 2022, scale = 4500, tessellation = "kunta")
print(munis)

Visualize the municipalities using ggplot2:

ggplot(munis) +
  geom_sf() +
  theme_minimal() +
  labs(title = "Finnish Municipalities (2022)")

Example: Spatial Filtering with a Bounding Box

To retrieve municipalities within a specific area (e.g., southern Finland), use the bbox parameter. Coordinates should match the specified crs.

bbox <- "200000,6600000,500000,6900000"  # In EPSG:3067
munis_south <- ogc_get_statfi_area(
  year = 2022,
  scale = 4500,
  tessellation = "kunta",
  bbox = bbox,
  crs = 3067
)

Visualize the filtered results:

ggplot(munis_south) +
  geom_sf() +
  theme_minimal() +
  labs(title = "Municipalities in Southern Finland (2022)")

Example: Fetching Wellbeing Areas

Retrieve wellbeing areas (hyvinvointialue) for 2022:

wellbeing <- ogc_get_statfi_area(
  year = 2022,
  tessellation = "hyvinvointialue",
  scale = 4500
)

Step 2: Retrieving Population Data by Administrative Area

The ogc_get_statfi_area_pop() function fetches administrative area polygons with associated population data, pivoted into a wide format where each population variable is a column. Parameters include:

year: The year of the data (2019–2021).
crs: Coordinate reference system (EPSG:3067 or EPSG:4326).
limit: Maximum number of features (or NULL for all).
bbox: Bounding box for spatial filtering.

Example: Fetching Population Data

Retrieve population data for 2021:

pop_data <- ogc_get_statfi_area_pop(year = 2021, crs = 3067)
print(pop_data)

Visualize population density (assuming a variable like population_total exists):

ggplot(pop_data) +
  geom_sf(aes(fill = population_total)) +
  scale_fill_viridis_c(option = "plasma") +
  theme_minimal() +
  labs(title = "Population by Administrative Area (2021)", fill = "Population")

Example: Population Data with Bounding Box

Fetch population data within a bounding box:

bbox <- "200000,6600000,500000,6900000"
pop_south <- ogc_get_statfi_area_pop(year = 2021, bbox = bbox, crs = 3067)

Step 3: Retrieving Population Data by Statistical Grid

The ogc_get_statfi_statistical_grid() function retrieves population data for statistical grid cells at 1km or 5km resolution. Data is returned in EPSG:3067 (ETRS89 / TM35FIN). Parameters include:

year: The year of the data (2019–2021).
resolution: Grid cell size (1000m or 5000m).
limit: Maximum number of features (or NULL for all).
bbox: Bounding box for spatial filtering.

Example: Fetching 5km Grid Data

Retrieve population data for a 5km grid in 2021:

grid_data <- ogc_get_statfi_statistical_grid(year = 2021, resolution = 5000)
print(grid_data)

Visualize the grid data:

ggplot(grid_data) +
  geom_sf(aes(fill = population_total), color = NA) +
  scale_fill_viridis_c(option = "magma") +
  theme_minimal() +
  labs(title = "Population by 5km Grid Cells (2021)", fill = "Population")

Example: 1km Grid with Bounding Box

Fetch 1km grid data within a bounding box:

bbox <- "200000,6600000,500000,6900000"
grid_south <- ogc_get_statfi_statistical_grid(
  year = 2021,
  resolution = 1000,
  bbox = bbox
)

Advanced Features

Pagination

When limit = NULL, the fetch_ogc_api_statfi() function automatically paginates through large datasets, fetching up to 10,000 features per request. This ensures all available data is retrieved, even for large administrative or grid datasets.

Error Handling

The package includes robust error handling:

Validates inputs (e.g., year, scale, tessellation, CRS, bounding box format).
Provides informative error messages for API failures or invalid responses.
Returns NULL with a warning if no data is retrieved, helping users diagnose issues.

Coordinate Reference Systems

The functions support two CRS options:

EPSG:3067 (ETRS89 / TM35FIN): The default for Finnish spatial data, suitable for local analyses.
EPSG:4326 (WGS84): Useful for global compatibility or web mapping.

Note that ogc_get_statfi_statistical_grid() is fixed to EPSG:3067, as per the API’s design.

Bounding Box Filtering

The bbox parameter allows spatial filtering to focus on specific regions. Ensure coordinates match the specified crs (e.g., EPSG:3067 for grid data). Example format: “200000,6600000,500000,6900000”.

Best Practices

Test with Limits: For large datasets (e.g., 1km grids), start with a small limit or bbox to estimate runtime before fetching all features.
CRS Selection: Use EPSG:3067 for Finnish data unless you need EPSG:4326 for compatibility with other systems.
Check Tessellation Types: Verify valid tessellation options (kunta, hyvinvointialue, etc.) when using ogc_get_statfi_area().
Inspect Output: Population data from ogc_get_statfi_area_pop() and ogc_get_statfi_statistical_grid() is pivoted into wide format. Check column names to identify available variables.