fix: ensure --crs-groups flag set for S2 conversions

The --crs-groups flag triggers prepare_dataset_with_crs_info() in data-model,
which writes CRS metadata via ds.rio.write_crs() and creates the spatial_ref
coordinate variable required by TiTiler validation.

Restores working configuration from commit 21ea009.

Author: wietzesuijker

scripts/augment_stac_item.py

@@ -66,9 +66,8 @@ def add_visualization(item: Item, raster_base: str, collection_id: str) -> None:
             _add_tile_links(item, base_url, query, "Sentinel-1 GRD VH")
 
     elif coll_lower.startswith(("sentinel-2", "sentinel2")):
-        # S2: Point to overview level 0 for quicklook TCI
-        # Use /r10m/0/tci path to access the overview array with spatial_ref
-        var_path = "/quality/l2a_quicklook/r10m/0/tci"
+        # S2: Use colon separator for TiTiler variable path
+        var_path = "/quality/l2a_quicklook/r10m:tci"
         query = (
             f"variables={urllib.parse.quote(var_path, safe='')}&bidx=1&bidx=2&bidx=3&assets=TCI_10m"
         )

scripts/create_geozarr_item.py

@@ -5,7 +5,6 @@
 
 import argparse
 import logging
-import re
 from typing import Any
 from urllib.parse import urlparse
 
@@ -31,30 +30,6 @@ def s3_to_https(s3_url: str, endpoint: str) -> str:
     return f"https://{bucket}.{host}/{path}"
 
 
-def normalize_r60m_href(href: str) -> str:
-    """Add /0/ subdirectory to r10m/r20m/r60m paths to match GeoZarr output structure.
-
-    GeoZarr conversion creates /0/ subdirectories (overview level 0) for all
-    resolution bands. This normalizes asset hrefs accordingly.
-
-    Example: .../r10m/tci → .../r10m/0/tci
-             .../r60m/b09 → .../r60m/0/b09
-    """
-    # Check for any resolution level pattern
-    for res in ["r10m", "r20m", "r60m"]:
-        if f"/{res}/" not in href:
-            continue
-
-        # If already has /0/ or other digit subdirectory, don't modify
-        if re.search(rf"/{res}/\d+/", href):
-            continue
-
-        # Insert /0/ after /{res}/
-        href = re.sub(rf"/({res})/", r"/\1/0/", href)
-
-    return href
-
-
 def find_source_zarr_base(source_item: dict) -> str | None:
     """Find the base Zarr URL from source item assets."""
     for asset in source_item.get("assets", {}).values():
@@ -112,9 +87,6 @@ def create_geozarr_item(
                     subpath = old_href[len(source_zarr_base) :]
                     new_href = output_zarr_base + subpath
 
-                    # Normalize r60m paths to include /0/ subdirectory (GeoZarr structure)
-                    new_href = normalize_r60m_href(new_href)
-
                     # Convert to https if needed
                     if new_href.startswith("s3://"):
                         new_href = s3_to_https(new_href, s3_endpoint)

scripts/get_zarr_url.py

@@ -0,0 +1,38 @@
+#!/usr/bin/env python3
+"""Extract Zarr URL from STAC item - standalone script for workflow templates."""
+
+import sys
+
+import httpx
+
+
+def get_zarr_url(stac_item_url: str) -> str:
+    """Get Zarr asset URL from STAC item."""
+    r = httpx.get(stac_item_url, timeout=30.0, follow_redirects=True)
+    r.raise_for_status()
+    assets = r.json().get("assets", {})
+
+    # Priority: product, zarr, then any .zarr asset
+    for key in ["product", "zarr"]:
+        if key in assets and (href := assets[key].get("href")):
+            return str(href)
+
+    # Fallback: any asset with .zarr in href
+    for asset in assets.values():
+        if ".zarr" in asset.get("href", ""):
+            return str(asset["href"])
+
+    raise RuntimeError("No Zarr asset found in STAC item")
+
+
+if __name__ == "__main__":
+    if len(sys.argv) != 2:
+        print("Usage: get_zarr_url.py <stac_item_url>", file=sys.stderr)
+        sys.exit(1)
+
+    try:
+        zarr_url = get_zarr_url(sys.argv[1])
+        print(zarr_url)
+    except Exception as e:
+        print(f"Error: {e}", file=sys.stderr)
+        sys.exit(1)