Add ogc.osc.prov-processing-step: PROV profile for a single geospatial processing step

.gitignore

@@ -2,3 +2,7 @@ build-local/
 .idea/
 .claude/
 .volumes
+*.bak
+.DS_Store
+build/
+.DS_Store

README.md

@@ -1,8 +1,9 @@
+Note: see issue #11 — EarthCODE example URLs currently 404 and block the register build.
 # Open Science Building Blocks 
 
 The repository documents a set of [OGC Building Blocks](https:blocks.ogc.org). for Open Science, which in turn leverage reusable ontologies, schemas, APIs etc. from published OGC standards.
 
-The set of building blocks and the interoperability specifications composed from them are visble here https://ogcincubator.github.io/bblocks-openscience/
+The set of building blocks and the interoperability specifications composed from them are visble here https://nsnarayanam.github.io/bblocks-openscience/
 
 An overview of the architecture and status :
 

_sources/prov-processing-step/bblock.json

@@ -0,0 +1,13 @@
+{
+  "name": "Geospatial Processing Step Provenance",
+  "abstract": "A profile of the PROV-O building block constrained to the description of a single geospatial processing step, with a required link to a registered process type.",
+  "status": "under-development",
+  "dateTimeAddition": "2026-08-05T00:00:00Z",
+  "itemClass": "schema",
+  "version": "0.1.0",
+  "dateOfLastChange": "2026-08-05",
+  "maturity": "development",
+  "scope": "unstable",
+  "tags": ["provenance", "prov-o", "processing", "ospd"],
+  "dependsOn": ["ogc.ogc-utils.prov"]
+}

_sources/prov-processing-step/description.md

@@ -0,0 +1,20 @@
+# Geospatial Processing Step Provenance
+
+This building block profiles `ogc-utils.prov` down to a single unit: one
+geospatial processing step.
+
+PROV-O is deliberately general. Any valid PROV document is a valid description
+of a processing step, which makes validation weak and cross-platform comparison
+hard.
+
+OGC API - Processes Part 5 (draft 26-038) narrows this for job provenance: a
+job entity, a process entity, an activity joining them, and input/output
+artifacts carrying roles. This profile follows that shape.
+
+Part 5 does not define any link from a process to a registered process type.
+This profile makes that link required, via the processType property. It is the
+point at which the OSPD 2026 profile-plus-register pattern closes: the profile
+gives the structure, the register gives the controlled term.
+
+Under development as part of OSPD 2026, deliverable D104 (Aganitha Space
+Technologies, Workflow Profiler).

_sources/prov-processing-step/examples.yaml

@@ -0,0 +1,294 @@
+- title: Ingest of mandal-level vegetation and soil moisture
+  description: >-
+    Retrieval of source observations into the workflow. Ingest is where
+    external provenance ends and workflow provenance begins: the generated
+    entity is the first thing downstream steps can cite.
+  snippets:
+    - language: json
+      code: |
+        {
+          "id": "urn:aganitha:step:ingest-dicra:20260807T060000Z",
+          "provType": "Activity",
+          "processType": "https://example.org/ospd/process-types/ingest",
+          "label": "Ingest mandal NDVI and soil moisture",
+          "startedAtTime": "2026-08-07T06:00:00Z",
+          "endedAtTime": "2026-08-07T06:03:41Z",
+          "used": ["urn:aganitha:source:dicra-mandal-archive"],
+          "generated": ["urn:aganitha:dataset:mandal-ndvi-sm-merged"],
+          "qualifiedUsage": [
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:source:dicra-mandal-archive",
+              "hadRole": "https://example.org/ospd/roles/sourceArchive"
+            }
+          ]
+        }
+
+- title: Normalisation of vegetation index to a condition index
+  description: >-
+    Rescaling an observed index against its own historical extremes. The
+    reference extrema are inputs rather than inline constants, so the period
+    they were derived from is recoverable from the provenance.
+  snippets:
+    - language: json
+      code: |
+        {
+          "id": "urn:aganitha:step:vci:20260807T061000Z",
+          "provType": "Activity",
+          "processType": "https://example.org/ospd/process-types/index-normalisation",
+          "label": "Compute vegetation condition index from NDVI",
+          "used": [
+            "urn:aganitha:dataset:mandal-ndvi-sm-merged",
+            "urn:aganitha:reference:ndvi-extrema"
+          ],
+          "generated": ["urn:aganitha:dataset:vci"],
+          "qualifiedUsage": [
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:dataset:mandal-ndvi-sm-merged",
+              "hadRole": "https://example.org/ospd/roles/observedIndex"
+            },
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:reference:ndvi-extrema",
+              "hadRole": "https://example.org/ospd/roles/normalisationReference"
+            }
+          ]
+        }
+
+- title: Normalisation of soil moisture to a deficit index
+  description: >-
+    The companion to the vegetation index step, and the piece that closes
+    the gap in the chain: without it, the combined index consumes an entity
+    no recorded activity produced, so a consumer walking the provenance
+    backwards hits a dead end.
+  snippets:
+    - language: json
+      code: |
+        {
+          "id": "urn:aganitha:step:smdi:20260807T061500Z",
+          "provType": "Activity",
+          "processType": "https://example.org/ospd/process-types/index-normalisation",
+          "label": "Compute soil moisture deficit index",
+          "used": [
+            "urn:aganitha:dataset:mandal-ndvi-sm-merged",
+            "urn:aganitha:reference:soil-moisture-extrema"
+          ],
+          "generated": ["urn:aganitha:dataset:smdi"],
+          "qualifiedUsage": [
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:dataset:mandal-ndvi-sm-merged",
+              "hadRole": "https://example.org/ospd/roles/observedIndex"
+            },
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:reference:soil-moisture-extrema",
+              "hadRole": "https://example.org/ospd/roles/normalisationReference"
+            }
+          ]
+        }
+
+- title: Weighted composition of a combined drought index
+  description: >-
+    Combination of two condition indices under fixed weights. Both inputs
+    share the same role, and the weights are step parameters rather than
+    properties of the process type — a case that bears directly on how
+    finely process types should be registered.
+  snippets:
+    - language: json
+      code: |
+        {
+          "id": "urn:aganitha:step:cdsi:20260807T062000Z",
+          "provType": "Activity",
+          "processType": "https://example.org/ospd/process-types/index-composition",
+          "label": "Compose combined drought severity index",
+          "used": [
+            "urn:aganitha:dataset:vci",
+            "urn:aganitha:dataset:smdi"
+          ],
+          "generated": ["urn:aganitha:dataset:cdsi"],
+          "qualifiedUsage": [
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:dataset:vci",
+              "hadRole": "https://example.org/ospd/roles/indexComponent"
+            },
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:dataset:smdi",
+              "hadRole": "https://example.org/ospd/roles/indexComponent"
+            }
+          ],
+          "parameters": {
+            "weights": { "vci": 0.6, "smdi": 0.4 }
+          }
+        }
+
+- title: Classification into drought severity classes
+  description: >-
+    Assignment of continuous index values to discrete classes. The class
+    scheme is an input entity, so the thresholds applied are recoverable
+    from the provenance without reading the code.
+  snippets:
+    - language: json
+      code: |
+        {
+          "id": "urn:aganitha:step:classify:20260807T063000Z",
+          "provType": "Activity",
+          "processType": "https://example.org/ospd/process-types/classification",
+          "label": "Classify drought severity",
+          "used": [
+            "urn:aganitha:dataset:cdsi",
+            "urn:aganitha:scheme:drought-severity-classes"
+          ],
+          "generated": ["urn:aganitha:dataset:drought-classes"],
+          "qualifiedUsage": [
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:dataset:cdsi",
+              "hadRole": "https://example.org/ospd/roles/inputField"
+            },
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:scheme:drought-severity-classes",
+              "hadRole": "https://example.org/ospd/roles/classificationScheme"
+            }
+          ],
+          "parameters": {
+            "classBreaks": [0.30, 0.45, 0.60, 0.75]
+          }
+        }
+
+- title: Standardised precipitation index over an accumulation window
+  description: >-
+    Derivation of a standardised index from an accumulated series against a
+    fitted reference period. Two entities of the same kind distinguished
+    only by role — the case where entity references alone would be
+    ambiguous and qualifiedUsage does real work.
+  snippets:
+    - language: json
+      code: |
+        {
+          "id": "urn:aganitha:step:spi:20260807T064000Z",
+          "provType": "Activity",
+          "processType": "https://example.org/ospd/process-types/standardised-index",
+          "label": "Compute standardised precipitation index",
+          "used": [
+            "urn:aganitha:dataset:precipitation-series",
+            "urn:aganitha:dataset:precipitation-reference-period"
+          ],
+          "generated": ["urn:aganitha:dataset:spi"],
+          "qualifiedUsage": [
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:dataset:precipitation-series",
+              "hadRole": "https://example.org/ospd/roles/inputSeries"
+            },
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:dataset:precipitation-reference-period",
+              "hadRole": "https://example.org/ospd/roles/fittingReference"
+            }
+          ]
+        }
+
+- title: Model training as a processing step
+  description: >-
+    Training a predictive model, expressed as an ordinary processing
+    activity. The generated entity is a fitted model rather than a data
+    product, and the fold assignment is an input entity so the spatial
+    cross-validation design is recoverable from the provenance.
+
+    This is the case the profile is least obviously shaped for, and the
+    reason it is included: if a trained model cannot be expressed as an
+    entity generated by an activity, the profile cannot describe machine
+    learning workflows at all.
+  snippets:
+    - language: json
+      code: |
+        {
+          "id": "urn:aganitha:step:train:20260807T070000Z",
+          "provType": "Activity",
+          "processType": "https://example.org/ospd/process-types/model-training",
+          "label": "Train drought prediction model with spatial cross-validation",
+          "used": [
+            "urn:aganitha:dataset:cdsi",
+            "urn:aganitha:dataset:spi",
+            "urn:aganitha:dataset:drought-classes",
+            "urn:aganitha:partition:district-folds"
+          ],
+          "generated": [
+            "urn:aganitha:model:drought-classifier",
+            "urn:aganitha:report:validation-metrics"
+          ],
+          "qualifiedUsage": [
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:dataset:cdsi",
+              "hadRole": "https://example.org/ospd/roles/trainingFeatures"
+            },
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:dataset:spi",
+              "hadRole": "https://example.org/ospd/roles/trainingFeatures"
+            },
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:dataset:drought-classes",
+              "hadRole": "https://example.org/ospd/roles/trainingTarget"
+            },
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:partition:district-folds",
+              "hadRole": "https://example.org/ospd/roles/validationPartition"
+            }
+          ]
+        }
+
+- title: Inference using a trained model
+  description: >-
+    Applying a fitted model to produce a projection. The same model entity
+    that was generated by training is here consumed as an input — the
+    round trip that makes model provenance chainable, and something no
+    purely data-oriented profile has to handle.
+  snippets:
+    - language: json
+      code: |
+        {
+          "id": "urn:aganitha:step:project:20260807T071500Z",
+          "provType": "Activity",
+          "processType": "https://example.org/ospd/process-types/model-inference",
+          "label": "Generate forward drought projection",
+          "used": [
+            "urn:aganitha:model:drought-classifier",
+            "urn:aganitha:dataset:cdsi"
+          ],
+          "generated": ["urn:aganitha:dataset:drought-projection"],
+          "qualifiedUsage": [
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:model:drought-classifier",
+              "hadRole": "https://example.org/ospd/roles/fittedModel"
+            },
+            {
+              "@type": "prov:Usage",
+              "entity": "urn:aganitha:dataset:cdsi",
+              "hadRole": "https://example.org/ospd/roles/inferenceInput"
+            }
+          ]
+        }
+
+- title: Minimal conforming step
+  description: >-
+    The smallest document that satisfies this profile.
+  snippets:
+    - language: json
+      code: |
+        {
+          "id": "urn:example:step:0001",
+          "provType": "Activity",
+          "processType": "https://example.org/ospd/process-types/ingest",
+          "used": ["urn:example:input:0001"],
+          "generated": ["urn:example:output:0001"]
+        }