eurodatacube / public-collections Goto Github PK

The license link for the cop-dem is broken

We currently link the correct license for Sentinel collections indirectly

    {
      "href": "https://docs.sentinel-hub.com/api/latest/data/sentinel-1-grd/#attribution-and-use",
      "rel": "license",
      "type": "text/html",
      "title": "License"
    }

We got the request to link to the correct document directly. We should update the license for all Sentinel collections (S1-GRD, S2-L1C, S2-L2A, 3-SLSTR, 3-OLCI, S5P).

    {
      "href": "https://sentinels.copernicus.eu/documents/247904/690755/Sentinel_Data_Legal_Notice",
      "rel": "license",
      "type": "application/pdf",
      "title": "License"
    }

@jdries FYI

As discussed here, data type values should be converted from the current uppercase to lowercase, in order to comply with https://github.com/stac-extensions/raster#data-types.

Came across this collection which throws an error because there is a space in the band name

• Global Surface Water
• GHS Built S2

The search seems to be entirely broken currently and returning way to many results.

The population density collection seems to have the wrong temporal metadata leading to no visualization in the EDC Browser:

• https://github.com/eurodatacube/public-collections/blob/main/collections/population_density.yaml

The collection should in the end look very similar to how Sentinel/Landsat looks:

• Documentation
• Scripts used in href within the Configurations have to be added first to https://github.com/sentinel-hub/custom-scripts repository

Hi @eorlitova
I noticed a few things that needs to be changed. Please make the changes and assign the pull request to me

apologies for taking you back

1. YAML file

The page is rendered here : https://collections.eurodatacube.com/ghsl-urban-delineation/

Sorry I missed this in the review;

• Attribution
  License: "[Creative Commons Attribution 4.0 International License](https://creativecommons.org/licenses/by/4.0/). Attribution: Contains modified Copernicus Sentinel data (year) processed by Sentinel Hub"

I believe it is not processed by Sentinel Hub, is it ? if not please remove the attribution part

• Link to pricing info is not found
  Pricing: Check [here](ghsl-urban-delineation/PRICING.MD) for more info about pricing and restrictions.

Please reference directly including the root url like this: [here](https://collections.eurodatacube.com/ghsl-urban-delineation/pricing.html)

2. README file

The page is rendered here https://collections.eurodatacube.com/ghsl-urban-delineation/readme.html

• Pricing info not found
  Here again please reference directly including the root url like this: [Details in PRICING.MD](https://collections.eurodatacube.com/ghsl-urban-delineation/pricing.html)

• Result images not properly formated
  You need to introduce break lines or similar for better formating
  here is my example code suggestion that gives a better result

### First hierarchical level 

<img src="ghsl-urban-delineation-level1-nomenclature.png" alt="level1 momenclature" width="50%"/> 

<br>
<br>
First hierarchical level - definition

<br>
<img src="ghsl-urban-delineation-level1-definition.png" alt="level1 definition" width="50%"/>

<br>
<br>
First hierarchical level - result

<br>
<img src="ghsl-urban-delineation-level1-result.png" alt="polygons - result at the first hierarchical level" width="50%"/>
<br>
<br>

### Second hierarchical level 

<img src="ghsl-urban-delineation-level2-nomenclature.png" alt="level2 momenclature" width="50%"/>

<br>
<br>
Second hierarchical level - definition

<br>
<img src="ghsl-urban-delineation-level2-definition.png" alt="level2 definition" width="50%"/>  

<br>
<br>
Second hierarchical level - result

<br>
<img src="ghsl-urban-delineation-level2-result.png" alt="polygons - result at the second hierarchical level" width="50%"/>
<br>

Currently we create a LicenseUrl link object no matter weather LicenseUrl is defined in the yaml or not. This leads to an empty LicenseUrl as can be seen here: https://collections.eurodatacube.com/stac/tsmn_water_quality_saturday.json

We should only include this in the json if LicenseUrl is defined. Main difficulty here is to format the json correctly so that the comma before the LicenseUrl link object is placed if it's present and not if it's not.

Needs to be done here: https://github.com/eurodatacube/public-collections/blob/main/_build/stac.hbs

prbit should be orbit

Our https://collections.eurodatacube.com/stac/index.json endpoint generates for some collections empty IDs. We should either remove them from the endpoint or make sure they are generated correctly (with full metadata).

Collections of questions:

• https://collections.eurodatacube.com/stac/phytoplankton-primary-production.json
• https://collections.eurodatacube.com/stac/co_3daily_data.json
• https://collections.eurodatacube.com/stac/s5p-ch4-weekly.json
• https://collections.eurodatacube.com/stac/so2_daily_data.json

I've found the index of STAC collections here: https://collections.eurodatacube.com/stac/index.json Unfortunately it is not a valid STAC catalog and can therefore not be browsed with tools like STAC Browser.

A valid STAC catalog can be created with minimal effort, e.g.:

{
    "type": "Catalog",
    "stac_version": "1.0.0",
    "id": "eurodatacube-collections-catalog",
    "description": "A STAC catalog that contains all collections available on EuroDataCube.",
    "links": [
        {
            "href": "https://collections.eurodatacube.com/stac/index.json",
            "rel": "self"
        },
        {
            "href": "https://collections.eurodatacube.com/stac/index.json",
            "rel": "root"
        },
        {
            "id": "SENTINEL2_L2A_MOSAIC_120",
            "href": "https://collections.eurodatacube.com/stac/sentinel-s2-l2a-mosaic-120.json",
            "rel": "child"
        },
        ...
    ]
}

I think the relevant change would have to be made to this file:

Something along the lines of:

{
  "type": "Catalog",
  "stac_version": "1.0.0",
  "id": "eurodatacube-collections-catalog",
  "description": "A STAC catalog that contains all collections available on EuroDataCube.",
  "links": [
    {
      "href": "{{rootUrl}}stac/index.json",
      "rel": "self"
    },
    {
      "href": "{{rootUrl}}stac/index.json",
      "rel": "root"
    },
    {{#each datasets}}
      {"id": "{{OpenEOPID}}", "rel": "child", "href": "{{rootUrl}}stac/{{Slug}}.json"}{{#unless @last}},{{/unless}}
    {{/each}}
  ]
}

URL in the collection metadata is not using the correct new endpoint (Follow up from: #284)

URL the Script is stored but the custom script repository: https://custom-scripts.sentinel-hub.com/custom-scripts/copernicus_services/global-land-cover/global_land_cover/script.js

URL in the public collection metadata: https://custom-scripts.sentinel-hub.com/copernicus_services/global_land_cover/global-land-cover/script.js

• whenever this comes through as an option just make it available through the "?"
• https://sentinels.copernicus.eu/web/sentinel/data-product-quality-reports

Add raster:bands object to collection metadata summaries.

Example for S2-L2A B01 band

What we would need to add to the yaml files (for each band).

  raster:bands:
    - name: B01
      data_type: float32
      statistics:
        minimum: 
          - 0
        maximum:
          - 0.4

How it will look rendered as json:

"summaries": {
    "raster:bands": [
      {
        "name": "B01",
        "data_type": "float32",
        "statistics": {"minimum": 0, "maximum": 0.4}
      },

Related resources:

• EDC public collections repository has the metadata in text/table form for some collections (see the DocumentationLinks we added to the links object)
• info about most of the core SH collections can be found in our documentation e.g. here for S2-L2A
• Info on raster STAC extension used here: https://github.com/stac-extensions/raster
• related to this request within the openEO Platform project: https://github.com/openEOPlatform/architecture-docs/issues/178#issuecomment-1006606630

Sentinel-1 GRD and Sentinel-3 OLCI L1B collections contain additional available bands that are not added to the collections' metadata, although they are listed in the Documentation:

1. Missing bands for Sentinel-1 GRD collection
2. Missing band for Sentinel-3 OLCI L1B.

We currently list eo:cloud_cover as array in our collection metadata ("eo:cloud_cover":[0,100]). This should be rendered as range object in the json.

Task

Update the yaml file and modify stac.hbs to render eo:cloud_cover as range object.

This is how it should later look in the json:

"eo:cloud_cover": {
    "minimum": 0,
    "maximum": 100
},

Related platform issue: https://github.com/openEOPlatform/architecture-docs/issues/269