Publishing tiles to OGC API - Tiles

OGC API - Tiles provides access to geospatial data in the form of tiles (map, vector, coverage, etc.).

pygeoapi can publish tiles from local or remote data sources (including cloud object storage or a tile service).

Providers

pygeoapi core tile providers are listed below, along with supported features.

Provider

rendered on-the-fly

properties

WebMercatorQuad

WorldCRS84Quad

raster

vector

MVT-tippecanoe

❌

✅

✅

❌

❌

✅

MVT-elastic

✅

✅

✅

❌

❌

✅

MVT-proxy

❓

❓

❓

❓

❌

✅

WMTSFacade

✅

❌

✅

✅

✅

❌

MVT-postgresql

✅

✅

✅

✅

❌

✅

Below are specific connection examples based on supported providers.

Connection examples

MVT-tippecanoe

This provider gives support to serving tiles generated using Mapbox Tippecanoe. The tiles can be integrated from a path on disk, or from a static url (e.g.: from an S3 or MinIO bucket). In both cases, they have to be rendered before using pygeoapi.

This code block shows how to configure pygeoapi to read Mapbox vector tiles generated with tippecanoe, from disk or a URL.

providers:
    - type: tile
      name: MVT-tippecanoe
      data: tests/data/tiles/ne_110m_lakes  # local directory tree
      # data: http://localhost:9000/ne_110m_lakes/{z}/{x}/{y}.pbf # tiles stored on a MinIO bucket
      options:
          zoom:
              min: 0
              max: 5
     # MVT-tippecanoe always uses WebMercatorQuad tiling scheme
      format:
          name: pbf
          mimetype: application/vnd.mapbox-vector-tile

Tip

In the diving into pygeoapi workshop OGC API - Tiles Exercise, detailed instructions can be found on how to generate tiles using tippecanoe and integrate them into pygeoapi.

MVT-elastic

This provider gives support to serving tiles generated using Elasticsearch. These tiles are rendered on-the-fly using the Elasticsearch Vector tile search API. In order to use it, the only requirement is to have the data stored in an Elasticsearch index.

This code block shows how to configure pygeoapi to read Mapbox vector tiles from an Elasticsearch endpoint.

providers:
    - type: tile
      name: MVT-elastic
      data: http://localhost:9200/ne_110m_populated_places_simple2/_mvt/geometry/{z}/{x}/{y}?grid_precision=0
      # if you don't use precision 0, you will be requesting for aggregations which are not supported in the
      # free version of elastic
      options:
          zoom:
              min: 0
              max: 5
     # MVT-elastic always uses WebMercatorQuad tiling scheme
      format:
          name: pbf
          mimetype: application/vnd.mapbox-vector-tile

Tip

On this tutorial you can find detailed instructions on publish tiles stored in an Elasticsearch endpoint.

MVT-proxy

This provider gives support to serving tiles from a generic tiles provider {z}/{x}/{y}.

For example, you can get and publish tiles from PostGIS providers like pg_tileserver or martin. Both of them render tiles on the fly and provide properties.

Following block shows how to configure pygeoapi to read Mapbox vector tiles from pg_tileserver endpoint.

providers:
    - type: tile
      name: MVT-proxy
      data: http://localhost:7800/public.ne_50m_admin_0_countries/{z}/{x}/{y}.mvt
      options:
          zoom:
              min: 0
              max: 15
          schemes:
              - WebMercatorQuad # this option is needed in the MVT-proxy provider
      format:
          name: pbf
          mimetype: application/vnd.mapbox-vector-tile

Following code block shows how to configure pygeoapi to read Mapbox vector tiles from martin endpoint.

providers:
    - type: tile
      name: MVT-proxy
      data: http://localhost:3000/ne_50m_admin_0_countries/{z}/{x}/{y}
      options:
          zoom:
              min: 0
              max: 15
          schemes:
             - WebMercatorQuad
      format:
          name: pbf
          mimetype: application/vnd.mapbox-vector-tile

MVT-postgresql

Note

Requires Python packages sqlalchemy, geoalchemy2 and psycopg2-binary

Note

Must have PostGIS installed with protobuf-c support

This provider gives support to serving tiles generated using PostgreSQL with PostGIS. The tiles are rendered on-the-fly using ST_AsMVT and related methods.

This code block shows how to configure pygeoapi to render Mapbox vector tiles from a PostGIS table.

providers:
    - type: tile
      name: MVT-postgresql
      data:
          host: 127.0.0.1
          port: 3010 # Default 5432 if not provided
          dbname: test
          user: postgres
          password: postgres
          search_path: [osm, public]
      id_field: osm_id
      table: hotosm_bdi_waterways
      geom_field: foo_geom
      storage_crs: http://www.opengis.net/def/crs/EPSG/0/4326
      options:
          zoom:
              min: 0
              max: 15
      format:
          name: pbf
          mimetype: application/vnd.mapbox-vector-tile

Tip

Geometry must have correctly defined storage_crs

PostgreSQL-related connection options can also be added to options. Please refer to the PostgreSQL OGC Features Provider documentation for more information.

WMTSFacade

This provider gives support to serving map tiles generated using a WMTS.

It is important that the WMTS provides a tile matrix set that matches exactly one of the configured schemes in pygeoapi. Currently only WebMercatorQuad and WorldCRS84Quad are available in pygeopi.

This code block shows how to configure pygeoapi to read map tiles from a WMTS.

providers:
    - type: tile
      name: WMTSFacade
      data: https://emotional.byteroad.net/geoserver/gwc/service/wmts
      format:
          name: png  # png or jpeg
          mimetype: image/png
      options:
          wmts_layer: camb:hex350_grid_mental_1920 # the layer name of the wmts
          wmts_tile_matrix_set: WebMercatorQuad  # the name of the tile matrix set of the wmts.
          wmts_style: camb:hex350_grid_mental_1920  # the style identifier of the wmts. If empty or this key is missing, it falls back to the default style.
          scheme: WebMercatorQuad  # the aligning scheme in pygeoapi.
          zoom:
              min: 0
              max: 20

Providing custom Tile Matrix Set definitions

By default, pygeoapi provides the WorldCRS84Quad and WebMercatorQuad TMS definitions, for tile providers to use accordingly. Additional TMS definitions may be added in pygeoapi’s resources/definitions/tiles (for example, by adding TMS definition files directly, volume mapping / Docker COPY, Docker Compose volumes, etc.).

Data access examples