Skip to content

Exercise 4 - Tiles of geospatial data via OGC API - Tiles

OGC API - Tiles provides a Web API to deliver tiles of geospatial information. Different forms of geospatial information are supported, such as tiles of vector features ("vector tiles"), coverages, maps (or imagery) and potentially eventually additional types of tiles of geospatial information. The standard is available on this document:

Note

OGC API - Tiles extends the collections/* URL structure (tilesets are listed under /collections/example/tiles:

https://demo.pygeoapi.io/collections/lakes/tiles/WebMercatorQuad/{tileMatrix}/{tileRow}/{tileCol}?f=mvt

pygeoapi support

pygeoapi supports the core OGC API - Tiles specification, and is able to advertise an existing tileset. Note that pygeoapi itself does not render tiles from source data, but it supports publishing tiles from different backend providers.

Note

The OGC API - Tiles URL structure is compatible with XYZ layers in common libraries such as OpenLayers, Leaflet and MapML

Note

See the official documentation for more information on supported tile backends

Note

pygeoapi currently supports two well known Tile Matrix Sets: WorldCRS84Quad and WebMercatorQuad. Their definition is published on the /TileMatrixSets end point.

Publish pre-rendered vector tiles

In this scenario, tiles must be pre-rendered before serving. The MVT-tippecanoe provider enables serving tiles pre-rendered by tippecanoe, either from a path on disk or from a static server (e.g.: S3 or MinIO bucket).

For this exercise, you will publish a vector dataset of the greater Hyderabad municipal corporation ward boundaries, from the location below:

  • data: workshop/exercises/data/hyderabad/greater_hyderabad_municipal_corporation_ward_Boundaries.geojson

Let's generate the tiles as the first step using tippecanoe:

Using tippecanoe to generate vector tiles

cd workshop/exercises
docker run -it --rm -v $(pwd)/data:/data emotionalcities/tippecanoe \
tippecanoe -r1 -pk -pf --output-to-directory=/data/tiles/ --force --maximum-zoom=16 \
--extend-zooms-if-still-dropping --no-tile-compression /data/hyderabad/greater_hyderabad_municipal_corporation_ward_Boundaries.geojson


cd workshop/exercises
docker run -it --rm -v ${pwd}/data:/data emotionalcities/tippecanoe tippecanoe -r1 -pk -pf --output-to-directory=/data/tiles/ --force --maximum-zoom=16 --extend-zooms-if-still-dropping --no-tile-compression /data/hyderabad/greater_hyderabad_municipal_corporation_ward_Boundaries.geojson

Note

Please note that the tippecanoe tool requires the input file to be in WGS84, and it always outputs tiles in WebMercator.

Update the pygeoapi configuration

Open the pygeoapi configuration in a text editor. Add a new dataset section as follows:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
    hyderabad:
        type: collection
        title: Greater Hyderabad Municipal Corporation ward boundaries
        description: The city ward boundaries represent the administrative and electoral boundary areas of the city. It plays a great role in planning of the city, for each council of the municipal corporation.
        keywords:
           - Boundaries
           - Administrative
           - Ward
        links:
            - type: text/html
              rel: canonical
              title: information
              href: https://livingatlas-dcdev.opendata.arcgis.com/datasets/a090c89d52f1498f96a82e97b8bfb83e_0/about
              hreflang: en-US
        extents:
            spatial:
                bbox: [78.2379194985166180,17.2908061510471995,78.6217049083810764,17.5618443356918768]
                crs: http://www.opengis.net/def/crs/OGC/1.3/CRS84
            temporal:
                begin: 2011-11-11
                end: null  # or empty
        providers:
            - type: feature
              name: GeoJSON
              data: /data/hyderabad/greater_hyderabad_municipal_corporation_ward_Boundaries.geojson
              id_field: objectid
            - type: tile
              name: MVT-tippecanoe
              data: /data/tiles/  # local directory tree
              options:
                zoom:
                    min: 0
                    max: 16
              format:
                    name: pbf
                    mimetype: application/vnd.mapbox-vector-tile

Save the file and restart Docker Compose. Navigate to http://localhost:5000/collections to evaluate whether the new dataset has been published.

Additional check for the following tile specific endpoints in the hyderabad collection:

TileSet

Publish vector tiles from Elasticsearch

Elasticsearch provides a middleware that renders an index on the fly, as vector tiles. This middleware is also supported by the pygeoapi mvt backend.

If you want to explore publishing vector tiles using Elasticsearch clone pygeoapi-examples repository:

git clone https://github.com/geopython/pygeoapi-examples.git


git clone https://github.com/geopython/pygeoapi-examples.git

Then change into the docker/mvt-elastic folder:

cd docker/mvt-elastic


cd docker/mvt-elastic

Edit the add-data.sh script on the ES folder, adding these two lines before the end:

1
2
    curl -o /tmp/hyderabad.geojson https://raw.githubusercontent.com/geopython/diving-into-pygeoapi/refs/heads/main/workshop/exercises/data/hyderabad/greater_hyderabad_municipal_corporation_ward_Boundaries.geojson
    python3 /load_es_data.py /tmp/hyderabad.geojson objectid

Above we are downloading the greater_hyderabad_municipal_corporation_ward_Boundaries.geojson inside the container, and ingesting it into an Elasticsearch index. After this we need to build the docker image:

docker compose build


docker compose build

Edit the docker.config.yml configuration on the pygeoapi folder, adding this code block before the end:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
    hyderabad:
        type: collection
        title: Greater Hyderabad Municipal Corporation ward boundaries
        description: The city ward boundaries represent the administrative and electoral boundary areas of the city. It plays a great role in planning of the city, for each council of the municipal corporation.
        keywords:
           - Boundaries
           - Administrative
           - Ward
        links:
            - type: text/html
              rel: canonical
              title: information
              href: https://livingatlas-dcdev.opendata.arcgis.com/datasets/a090c89d52f1498f96a82e97b8bfb83e_0/about
              hreflang: en-US
        extents:
            spatial:
                bbox: [78.2379194985166180,17.2908061510471995,78.6217049083810764,17.5618443356918768]
                crs: http://www.opengis.net/def/crs/OGC/1.3/CRS84
            temporal:
                begin: 2011-11-11
                end: null  # or empty
        providers:
            - type: feature
              name: Elasticsearch
              #Note elastic_search is the docker container of ES the name is defined in the docker-compose.yml
              data: http://elastic_search:9200/hyderabad
              id_field: objectid
            - type: tile
              name: MVT-elastic
              data: http://elastic_search:9200/hyderabad/_mvt/geometry/{z}/{x}/{y}?grid_precision=0
              # index must have a geo_point
              options:
                zoom:
                    min: 0
                    max: 29
              format:
                    name: pbf
                    mimetype: application/vnd.mapbox-vector-tile

This configuration enables publishing greater_hyderabad_municipal_corporation_ward_Boundaries.geojson as both, OGC API - Features and OGC API - Tiles.

Note

The elastic Vector tile search API supports zoom levels 0-29

Finally start the docker composition, which will download and ingest the dataset and publish it in pygeoapi:

docker compose up


docker compose up

Note

You can check your elastic index at: http://localhost:9200/_cat/indices

If you are in production, you may want to close the elastic ports on docker-compose.

Client access

LeafletJS

LeafletJS is a popular JavaScript library to add interactive maps to websites. LeafletJS does not support OGC API's explicitly, however can interact with OGC API by using the results of the API directly.

Add OGC API - Tiles to a website with LeafletJS

  • copy the HTML below to a file called vector-tiles.html, or locate this file in workshop/exercises/html
  • open the file in a web browser

The code uses the LeafletJS library with the leaflet.vectorgrid plugin to display the lakes OGC API - Tiles service on top of a base layer.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
<html>
<head><title>OGC API - Tiles exercise</title></head>
<body>
<div id="map" style="width:100vw;height:100vh;"></div>
<link rel="stylesheet" href="https://unpkg.com/leaflet@1.0.3/dist/leaflet.css" />
<script type="text/javascript" src="https://unpkg.com/leaflet@1.3.1/dist/leaflet.js"></script>
<script type="text/javascript" src="https://unpkg.com/leaflet.vectorgrid@1.3.0/dist/Leaflet.VectorGrid.bundled.js"></script>
<script>    
map = L.map('map').setView({ lat: 17.425181, lng: 78.5493906 }, 11);
map.addLayer(
    new L.tileLayer('https://server.arcgisonline.com/ArcGIS/rest/services/NatGeo_World_Map/MapServer/tile/{z}/{y}/{x}', {
    attribution: 'Tiles &copy; Esri &mdash; National Geographic, Esri, DeLorme, NAVTEQ, UNEP-WCMC, USGS, NASA, ESA, METI, NRCAN, GEBCO, NOAA, iPC',
    minZoom: 1,
    maxZoom: 16,
    }));
 function getColor(val){
    if (val < 40) {return "#f2e6c7"}
    else if (val < 80) {return "#8fa37e"}
    else if (val < 100) {return "#f0d17d"}
    else if (val < 120) {return  "#d7ded1"}
    else return "#c2d0d9";
 }
 var vectorTileStyling = {
    greater_hyderabad_municipal_corporation_ward_Boundaries: function(properties) {
        return ({
            fill: true,
            fillColor: getColor(properties.objectid),
            color: "#ffffff",
            fillOpacity: 1.0,
            weight: 5,
            //color: "#ffffff",
            opacity: 1.0,
        });
    }
} 
    var mapVectorTileOptions = {
        rendererFactory: L.canvas.tile,
        interactive: true,
        vectorTileLayerStyles: vectorTileStyling,
        };
    var pbfURL='http://localhost:5000/collections/hyderabad/tiles/WorldCRS84Quad/{z}/{x}/{y}?f=mvt';
    var pbfLayer=L.vectorGrid.protobuf(pbfURL,mapVectorTileOptions).on('click',function(e) {
        console.log(e.layer);
    L.DomEvent.stop(e);
    }).addTo(map); 
</script>
</body>
</html>

In this example, the colors of the symbols reflect the value of the objectid attribute.

Note

You can check the layer attributes, by opening the console in the developer tools.

Tip

Try adding a different pygeoapi vector tiles layer by updating the code in workshop/exercises/html/vector-tiles.html.

If you want to render the tiles from the Elasticsearch example, you can check out the code from this repository:

git clone -b ogcapi-ws https://github.com/emotional-cities/vtiles-example.git


git clone -b ogcapi-ws https://github.com/emotional-cities/vtiles-example.git

Tip

See the official LeafletJS documentation

OpenLayers

OpenLayers is a popular JavaScript library to add interactive maps to websites. OpenLayers natively supports OGC API - Tiles (map and vector). You can check the code for the example bellow on: https://ogcincubator.github.io/ogcapi-tiles-demo/

Tip

See the official OpenLayers documentation

Summary

Congratulations! You are now able to publish tiles to pygeoapi. You can learn more about this standard on: https://tiles.developer.ogc.org/