Ortho photo coverage in Lower Saxony

About LGLN opendata

The State Office for Geoinformation and Land Surveying Lower Saxony (LGLN) provides a lot of its geodata as open data. And since some years they provide ortho images as well as terrain models of Lower Saxony as Cloud Optimized GeoTIFFs (COG) and reference them via Spatio Temporal Asset Catalogs (STAC). Here we access the data via R to explore, vizualize and work with it programatically.

More information under opengeodata.lgln.niedersachsen.de.

Working with Orthoimagery

Initialize metadata catalog

The index of available data is available as a geojson file which is stored in a cloud bucket. It consists of many polygons which cover Lower Saxony and represent image tiles (2x2 km). For each tile it contains some basic metadata and links to further metadata of the images and links to the image data itself.

# read geojson from cloud bucket
dat <- sf::st_read("https://arcgis-geojson.s3.eu-de.cloud-object-storage.appdomain.cloud/dop20/lgln-opengeodata-dop20.geojson")

# modify date
dat <- dat |> 
  dplyr::mutate(
    date = lubridate::ymd(Aktualitaet),
    year = lubridate::year(date),
    month = lubridate::month(date, label = TRUE),
    doy = lubridate::yday(date)
  )

Plot metadata catalog

The data catalog contains information about the acquisition date of images. We can plot this data to get more information about the available images.

earliest / latest

# first and last available acquisition dates
dat |> arrange(date) |>  slice(1) |> pull(date)

dat |> arrange(desc(date)) |>  slice(1) |> pull(date)

Most recent coverage

# latest acquisition date per tile
dat |> 
  dplyr::group_by(tile_id) |> 
  dplyr::slice_max(date, n = 1) |> 
  ggplot2::ggplot() + 
    ggplot2::geom_sf(aes(fill = as.factor(year)), color = NA) + 
    ggplot2::theme_void()  +
    ggplot2::scale_fill_viridis_d(option = "plasma")

First coverage

# first acquisition date per tile
dat |> 
  dplyr::group_by(tile_id) |> 
  dplyr::slice_min(date, n = 1) |> 
  ggplot2::ggplot() + 
    ggplot2::geom_sf(aes(fill = as.factor(year)), color = NA) + 
    ggplot2::theme_void()  +
    ggplot2::scale_fill_viridis_d(option = "plasma")

Leaf-off coverage

# latest leaf-off data per tile
dat |>  
  dplyr::filter(month %in% c("Nov", "Dez", "Jan", "Feb", "Mrz")) |> 
  dplyr::arrange(date) |> 
  ggplot2::ggplot() + 
    ggplot2::geom_sf( aes(fill = as.factor(year)), color = NA) + 
    ggplot2::theme_void()  +
    ggplot2::scale_fill_viridis_d(option = "plasma")

Leaf-on coverage

# latest leaf-on data per tile
dat |>  
  dplyr::filter(month %in% c("Apr", "Mai", "Jun", "Jul", "Aug", "Sep", "Okt")) |> 
  dplyr::arrange(date) |> 
  ggplot2::ggplot() + 
    ggplot2::geom_sf( aes(fill = as.factor(year)), color = NA) + 
    ggplot2::theme_void()  +
    ggplot2::scale_fill_viridis_d(option = "plasma")

Acquisition month

# month of latest data per tile
dat |>  
  dplyr::group_by(tile_id) |> 
  dplyr::slice_max(date, n = 1) |> 
  ggplot2::ggplot() + 
    ggplot2::geom_sf( aes(fill = month), color = NA) + 
    ggplot2::theme_void()  +
    ggplot2::scale_fill_viridis_d(option = "plasma")