← Back to catalogue
Published

Weather

vr.wm-ear-001 · wm-ear-001-weather

Describe a bounded collection of observed and predicted atmospheric-state information with traceable sampling, uncertainty and external warning references.

World Models Physical world and living systems PHY.EAR.WTH

Bundle → Layer → Finding → Questions Filled

6 bundles · 12 layers · 12 findings · 48 questions

Weather context Bound the information collection and preserve its distinct information kinds.

Collection identity

A proposed weather aggregate identifies a bounded collection of atmospheric assertions, not a station, the atmosphere itself or a single storm. Members retain independent identities; empty declared coverage is not evidence of fair weather.

Collection identity

A proposed weather aggregate identifies a bounded collection of atmospheric assertions, not a station, the atmosphere itself or a single storm. Members retain independent identities; empty declared coverage is not evidence of fair weather.

  1. Which master identifier and revision identify this weather collection? identity
  2. What place and time window bound the collection membership? spatial
  3. Which independently identified observations, fields or products are members? composition
  4. Which accountable role stewards the collection and its usage policy? ownership

Information kind

Observation, retrieval, analysis, forecast and warning are distinguishable assertions. Weather data may feed a climate product without acquiring climate ownership.

Information kind

Observation, retrieval, analysis, forecast and warning are distinguishable assertions. Weather data may feed a climate product without acquiring climate ownership.

  1. Is this member an observation, retrieval, analysis, forecast or warning reference? classification
  2. Which atmospheric quantity or weather phenomenon does the assertion describe? definition
  3. Which climate or neighboring subject consumes or contextualizes this weather assertion? relationship
  4. What prevents a forecast or climate normal from being presented as an observed value? constraint
Observations and measurement context Preserve what was sensed and how to interpret its sampling.

Measured result

A meteorological observation carries a value or explicit unavailable state, a defined variable and its sampling support. The proposed fields accommodate surface and upper-air records without making every variable mandatory.

Measured result

A meteorological observation carries a value or explicit unavailable state, a defined variable and its sampling support. The proposed fields accommodate surface and upper-air records without making every variable mandatory.

  1. What value, unit and missing or trace state were actually reported? measurement
  2. When did the phenomenon occur and when was its result produced or ingested? temporal
  3. What averaging or accumulation interval and vertical level qualify the measurement? measurement
  4. What uncertainty, detection limit and quality flag qualify the reported result? quality

Observation source context

A reference to a station, mobile platform, sensor and procedure provides context while their commissioning, maintenance and calibration execution remain externally mastered. Historical location must be resolved at observation time.

Observation source context

A reference to a station, mobile platform, sensor and procedure provides context while their commissioning, maintenance and calibration execution remain externally mastered. Historical location must be resolved at observation time.

  1. Which station or mobile platform and sensor produced this observation? relationship
  2. Where was the sampled feature relative to the observing platform at that time? spatial
  3. Which procedure and calibration evidence were applicable when the sample was taken? provenance
  4. Which exposure change, relocation or outage limits comparability or availability? exception
Derived atmospheric fields Describe grids, retrievals and analyses without conflating inference with direct sensing.

Field support

A field binds each variable to coordinate and support metadata. Grid spacing is not a claim of effective resolving power; domain edges and absent cells remain explicit.

Field support

A field binds each variable to coordinate and support metadata. Grid spacing is not a claim of effective resolving power; domain edges and absent cells remain explicit.

  1. Which coordinates, grid mapping and cell bounds locate each field value? spatial
  2. Which units, statistic and cell methods describe the field variable? measurement
  3. Which valid times and original calendar qualify the field coordinates? temporal
  4. Which masked cells, domain limits or sampling gaps must consumers preserve? exception

Retrieval and analysis lineage

A derived atmospheric estimate records its method and input dependencies. Radar, satellite and blended analyses do not become direct point observations; assimilation and retrieval are not assumed interchangeable.

Retrieval and analysis lineage

A derived atmospheric estimate records its method and input dependencies. Radar, satellite and blended analyses do not become direct point observations; assimilation and retrieval are not assumed interchangeable.

  1. Which algorithm version and input revisions produced this atmospheric field? provenance
  2. Was the estimate retrieved, assimilated, interpolated or otherwise transformed? classification
  3. What validation evidence and uncertainty limits apply to this derived estimate? quality
  4. Which external Earth-observation product or processing activity remains the source master? relationship
Forecast products and uncertainty Record prediction context without running a forecasting system.

Forecast issue and run

Forecast products retain source model or method, cycle, member role and distinct temporal axes. A nowcast or human forecast can lack a numerical model run and must state its alternative method.

Forecast issue and run

Forecast products retain source model or method, cycle, member role and distinct temporal axes. A nowcast or human forecast can lack a numerical model run and must state its alternative method.

  1. Which forecast product revision and generating run or method are referenced? identity
  2. What are the initialization, issue, valid interval and lead time of this forecast? temporal
  3. Is the forecast deterministic, a named member, an ensemble statistic or a nowcast? classification
  4. Which earlier product is superseded and what remains available as issued? lifecycle

Probabilistic interpretation

Probability refers to a defined event over specified support. Spread, mean and probability are different products; an ensemble average need not represent a physically realized member.

Probabilistic interpretation

Probability refers to a defined event over specified support. Spread, mean and probability are different products; an ensemble average need not represent a physically realized member.

  1. Which threshold, inequality, variable, area and interval define the forecast event? definition
  2. What probability or quantile and sample basis does the product report? measurement
  3. What calibration or skill evidence applies to this product and use context? quality
  4. What limitations prevent spread or ensemble mean from becoming a certainty claim? constraint
Weather warning references Preserve authoritative messages and their qualified applicability without issuing alerts.

Warning identity and authority

A local weather record may reference an external warning and its meteorological evidence. CAP-like syntax and attribution alone neither authorize the issuer nor prove a hazard occurred.

Warning identity and authority

A local weather record may reference an external warning and its meteorological evidence. CAP-like syntax and attribution alone neither authorize the issuer nor prove a hazard occurred.

  1. Which sender, message identifier and sent time identify the warning? identity
  2. What evidence establishes the issuer remit for this area and warning type? authority
  3. Is the message actual, draft, exercise or test and what is its message type? classification
  4. Which observations, forecasts or expert assessments support this warning reference? evidence

Warning area and change history

Warning applicability carries area, timing and source severity/certainty/urgency. Cancellation or expiry changes the message status; it does not prove atmospheric safety.

Warning area and change history

Warning applicability carries area, timing and source severity/certainty/urgency. Cancellation or expiry changes the message status; it does not prove atmospheric safety.

  1. Which source area geometry or geocode and altitude range define the warning coverage? spatial
  2. What effective, onset and expiry times are declared by the warning issuer? temporal
  3. What severity, urgency and certainty values does the source message assert? state
  4. Which referenced update or cancellation changes the message chain as of the requested time? event
Evidence continuity and exchange Keep corrections, distributions and fitness evidence traceable.

Quality review and correction

Review history retains original and corrected assertions with their evaluation basis. A flagged value may remain useful for a qualified purpose; a corrected value is not silently substituted into an as-issued forecast evaluation.

Quality review and correction

Review history retains original and corrected assertions with their evaluation basis. A flagged value may remain useful for a qualified purpose; a corrected value is not silently substituted into an as-issued forecast evaluation.

  1. Which quality procedure and rule version produced the current review disposition? validation
  2. Which earlier assertion is corrected, withdrawn or replaced and for what reason? lifecycle
  3. Which dependent local views need re-evaluation after a source correction? process
  4. Which retention rule permits preservation or disposal of each weather record revision? retention

Delivery and usage profile

Storage and exchange formats are projections. A profile records provenance, rights, validation and loss so that transport success is never confused with scientific validity or freshness.

Delivery and usage profile

Storage and exchange formats are projections. A profile records provenance, rights, validation and loss so that transport success is never confused with scientific validity or freshness.

  1. Which pinned exchange profile maps this weather record and what information is lost? interoperability
  2. Which audience and use policy permit reading or redistributing the source payload? access
  3. How fresh and complete is the feed relative to its declared update expectation? temporal
  4. Which structural, reference and semantic checks have evidence of passing? validation

Classifiers Filled

Family
World Models
Category
Physical world and living systems
Entry kind
aggregate
Navigation path
NAV.PHY.EAR.WTH
Domain
PHY.EAR.WTH
Industry
Cross-industry
Tags
weatherphy.ear.wth
Also called
P4

What it is Filled

A format-neutral weather information aggregate whose members retain independent source identity and revision. It records local interpretations of observations, atmospheric fields, forecast products and warning references for a declared domain and time scope. It does not own stations, climate archives, hazard response or numerical prediction systems.

In scope

  • Collection scope, typed members and atmospheric quantities with time, location, units and measurement context
  • Derived fields and forecast products with lineage, sampling support, uncertainty and revision context
  • Externally issued weather warning references and local evidence, correction and exchange governance

Out of scope

  • Climate normals, long-term trends/extremes, homogenization and climate prediction; these belong to WM-EAR-002
  • Station/platform equipment operations, calibration execution, raw EO processing, ocean state and air-quality compliance
  • Hazard impact/exposure, emergency decisions, warning issuance/dissemination and numerical weather prediction execution

Why it exists Filled

Describe a bounded collection of observed and predicted atmospheric-state information with traceable sampling, uncertainty and external warning references.

Distinguishing features Derived, awaiting review

  • Unlike WM-EAR-002: Proposed optional binding; target version and instance master must be pinned before operational use. Climate owns normals, long-term extremes/trends, homogenized climate series and climate prediction products. Weather supplies versioned inputs and references baselines without computing or governing them.
  • Unlike WM-EAR-012: Proposed optional binding; target version and instance master must be pinned before operational use. Environmental Monitoring Station owns station and deployment history. Weather binds time-qualified observation context; fixed stations are optional.
  • Unlike WM-EAR-009: Proposed optional binding; target version and instance master must be pinned before operational use. EO Product owns remotely sensed source products. Weather owns only local atmospheric interpretation and provenance references.
  • Unlike WM-EAR-005: Proposed optional binding; target version and instance master must be pinned before operational use. Natural Hazard owns occurrence, exposure, consequences and response context. A weather warning reference does not establish a hazard occurrence or trigger a response.
  • Unlike WM-EAR-007: Proposed optional binding; target version and instance master must be pinned before operational use. Air Quality owns pollutant assessment and health/compliance interpretation. Weather may reference atmospheric composition as a contextual input.
  • Unlike WM-EAR-010: Proposed optional binding; target version and instance master must be pinned before operational use. Ocean / Marine State owns ocean state. Marine weather may share platforms while maintaining the atmospheric variable boundary.

Note: Derived from boundary notes against neighbouring models.

What robots and AI may and may not do Derived, awaiting review

Must not

  • Use evidence-backed revisions with expected-version checks; never silently overwrite an issued forecast or external warning.
  • Deny ungranted writes and restricted-payload reads; release public projections only under an explicit source-use and audience policy.

May

  • Register bounded weather collection: Proposed, unimplemented local operation. Assign local collection identity and scope while retaining member identities.
  • Record qualified observation: Proposed, unimplemented local operation. Ingest a proposed observation record with value, unit, sampling and source context.
  • Link atmospheric field lineage: Proposed, unimplemented local operation. Record coordinate metadata and dependencies for a retrieved or analyzed field.
  • Record forecast interpretation: Proposed, unimplemented local operation. Bind product identity, temporal axes and optional probabilistic event semantics.
  • Link external weather warning: Proposed, unimplemented local operation. Preserve source warning identity, authority evidence and message chain.
  • Revise local assertion with evidence: Proposed, unimplemented local operation. Attach a correction and mark dependent local projections for review.

Note: Derived from functions, policies, CRUD and access rules; prohibitions were not authored for agents as such.

Moral aspects Missing, in the backlog

Not described yet. This gap is in the card backlog.

Owners Filled

Steward

Accountable weather-data steward role and delegation evidence

Roles

Weather-data steward
Approve collection scope, namespaces and profile ownership.
Observation curator
Record source context, units, missingness and qualified quality evidence.
Forecast reviewer
Check cycle, valid interval and uncertainty interpretation without issuing predictions.
Warning-reference reviewer
Verify authority evidence, source status and update-chain integrity without disseminating warnings.
Records and access custodian
Apply audience, licensing, retention and disposition policies.

Links to other meta-models Filled

references

  • WM-EAR-002 - Proposed optional binding; target version and instance master must be pinned before operational use. Climate owns normals, long-term extremes/trends, homogenized climate series and climate prediction products. Weather supplies versioned inputs and references baselines without computing or governing them.
  • WM-EAR-012 - Proposed optional binding; target version and instance master must be pinned before operational use. Environmental Monitoring Station owns station and deployment history. Weather binds time-qualified observation context; fixed stations are optional.
  • WM-EAR-009 - Proposed optional binding; target version and instance master must be pinned before operational use. EO Product owns remotely sensed source products. Weather owns only local atmospheric interpretation and provenance references.
  • WM-EAR-005 - Proposed optional binding; target version and instance master must be pinned before operational use. Natural Hazard owns occurrence, exposure, consequences and response context. A weather warning reference does not establish a hazard occurrence or trigger a response.
  • WM-EAR-007 - Proposed optional binding; target version and instance master must be pinned before operational use. Air Quality owns pollutant assessment and health/compliance interpretation. Weather may reference atmospheric composition as a contextual input.
  • WM-EAR-010 - Proposed optional binding; target version and instance master must be pinned before operational use. Ocean / Marine State owns ocean state. Marine weather may share platforms while maintaining the atmospheric variable boundary.

aligned

  • OGC SensorThings API 1.1 - Observation and source-context mapping only; no API conformance claimed.
  • CF Conventions 1.12 - Variable, grid, time and sampling metadata alignment; retain source calendar and quantify conversion loss.
  • CAP 1.2 - External weather warning message alignment only; no issuing or dissemination authority.
  • PROV-O 2013 - Derivation and revision alignment; evidence lineage is distinct from scientific correctness.

neighbor

  • WM-EAR-002 - Proposed optional binding; target version and instance master must be pinned before operational use. Climate owns normals, long-term extremes/trends, homogenized climate series and climate prediction products. Weather supplies versioned inputs and references baselines without computing or governing them.
  • WM-EAR-012 - Proposed optional binding; target version and instance master must be pinned before operational use. Environmental Monitoring Station owns station and deployment history. Weather binds time-qualified observation context; fixed stations are optional.
  • WM-EAR-009 - Proposed optional binding; target version and instance master must be pinned before operational use. EO Product owns remotely sensed source products. Weather owns only local atmospheric interpretation and provenance references.
  • WM-EAR-005 - Proposed optional binding; target version and instance master must be pinned before operational use. Natural Hazard owns occurrence, exposure, consequences and response context. A weather warning reference does not establish a hazard occurrence or trigger a response.
  • WM-EAR-007 - Proposed optional binding; target version and instance master must be pinned before operational use. Air Quality owns pollutant assessment and health/compliance interpretation. Weather may reference atmospheric composition as a contextual input.
  • WM-EAR-010 - Proposed optional binding; target version and instance master must be pinned before operational use. Ocean / Marine State owns ocean state. Marine weather may share platforms while maintaining the atmospheric variable boundary.

What else AI and robots need to interact with it Incomplete

Identity and identifiers required Filled

  • Authoritative master-system identifier with namespace
  • Governed global identifier or IRI
  • Adopting-Dimension UUID or ULID

Direct properties required Missing, in the backlog

Not described yet. This gap is in the card backlog.

Recognition required Missing, in the backlog

Not described yet. This gap is in the card backlog.

Capabilities and actions required Filled

  • Register bounded weather collection: Proposed, unimplemented local operation. Assign local collection identity and scope while retaining member identities.
  • Record qualified observation: Proposed, unimplemented local operation. Ingest a proposed observation record with value, unit, sampling and source context.
  • Link atmospheric field lineage: Proposed, unimplemented local operation. Record coordinate metadata and dependencies for a retrieved or analyzed field.
  • Record forecast interpretation: Proposed, unimplemented local operation. Bind product identity, temporal axes and optional probabilistic event semantics.
  • Link external weather warning: Proposed, unimplemented local operation. Preserve source warning identity, authority evidence and message chain.
  • Revise local assertion with evidence: Proposed, unimplemented local operation. Attach a correction and mark dependent local projections for review.

Hazards and failure modes required Missing, in the backlog

Not described yet. This gap is in the card backlog.

Standards and interfaces required Derived, awaiting review

  • OGC SensorThings API Part 1: Sensing Version 1.1
  • PROV-O: The PROV Ontology

Context of use required Filled

  • International conceptual alignments do not establish country-specific warning authority or permissions.
  • Weather/Climate is separated by subject and product purpose, not a universal day-count cutoff.

Sources Filled

  1. Climate versus Weather - National Weather Service
  2. Guide to Instruments and Methods of Observation (WMO-No. 8): publication and chapter index - World Meteorological Organization
  3. OGC SensorThings API Part 1: Sensing Version 1.1 - Open Geospatial Consortium
  4. NetCDF Climate and Forecast (CF) Metadata Conventions - CF Conventions community
  5. Forecast User Guide: Section 8.1.1 Basic ensemble products - European Centre for Medium-Range Weather Forecasts
  6. Common Alerting Protocol Version 1.2 - OASIS Open
  7. PROV-O: The PROV Ontology - World Wide Web Consortium

Open questions

  • Pin and review applicable WMO normative chapters and current adopted source revisions; complete coordinator HTTP and substantive source checks.
  • Develop adoption schemas and test missing versus zero, interval mismatch, mobile platforms, native calendars, stale forecasts, incomplete ensembles, out-of-order warning cancellation and lawful evidence disposal.
  • Validate candidate neighbor bindings and restore independent external review before any canonical or publishable-draft promotion.
  • Nested instance schemas, required-by-kind constraints, executable mappings and adversarial fixtures remain incomplete. Candidate fields are optional elicitation slots, not permission to accept incomplete operational records.
  • Numerical thresholds, observing tolerances, instrument classes, aviation/marine coding and forecast verification procedures require specialist profiles.
  • GRIB/BUFR, detailed WIGOS identifiers and metadata bindings, radar/satellite retrieval physics and assimilation schemas are not implemented.
  • Independent provider review and full source/version verification remain absent. WMO evidence is a topic index, not reviewed normative chapters.

Machine files

Provenance

world-models research · reviewable-draft

Built from: models/wm-ear-001-weather/spec.yaml