Streamlining STAC queries via pystac

[dmoglioni]

Optimizing STAC queries to expose services associated to a data collections (for example SH statistical API and the corresponding dataId).
Below an example @lubojr shared with me on how currently it is implemented (data collection: S5p-NO2).
To investigate how the steps below can be streamlined.
cc: @aapopescu

Step 1: Access the EODashboard STAC Catalog via PySTAC

from
pystac import
Catalog

URL to the root STAC catalog (adjust if necessary)

catalog_url
= "https://esa-eodashboards.github.io/eodashboard-catalog/trilateral/catalog.json"

Load the catalog

catalog
= Catalog.from_file(catalog_url)

Print basic information

print(f"Catalog ID:
{catalog.id}")

print(f"Description:
{catalog.description}")

indicators
= list(catalog.get_children())

print(f"Number of child indicators:
{len(indicators)}")

no2_indicators_s5p
= [

ind
for ind
in indicators

if
"NO2" in
ind.extra_fields.get("tags",{})

and
"Sentinel-5P" in
ind.extra_fields.get("satellite",{})

]

print("retrieved indicators",
no2_indicators_s5p)

we "know" that we need to take NO2_daily but metadata-wise it is tricky to deduct that from the collections

daily_no2_indicator
= next(ind
for
ind in
no2_indicators_s5p if
ind.id
== "NO2_daily")

Get child links that are STAC collections

child_collections
= []

for
link in
daily_no2_indicator.get_links("child"):

if
link.rel
== "child":

child_collection
= link.resolve_stac_object().target

child_collections.append(child_collection)

Show child collection IDs and titles

print(f"{len(child_collections)}
child collections found under 'no2_daily': indicator")

for
col in
child_collections:

print(f"-
{col.id}:
{col.title}")

get actual service links connected to this collection

daily_no2_collection
= next(ind
for
ind in
child_collections if
ind.id
== "NO2_daily")

for
link in
daily_no2_collection.links:

print("available link:",
link)

we need to extract the link with relationship "example"

example_link
= next(link
for
link in
daily_no2_collection.links if
link.rel ==
"example")

list available fields of the link:

print("properties of the example link",
example_link.extra_fields)

we see that the SH collection ID is in dataID field

print("retrieved collection id:",
example_link.extra_fields["dataId"])

[lubojr]

Sentinel Hub Statistics exposed info should be streamlined and moved on the Process level (connected to deprecating "services" in STAC).