Skip to main content

Boundaries

The au_company_climate_intelligence pack operates within defined data and methodology boundaries. Understanding these is important for assurance engagements and disclosure preparation.

Scope boundaries​

  • Scope 1 and 2: Full coverage via NGER corporate emissions and NGA factors
  • Scope 3: Reference factors plus indicative screening-grade estimates for Categories 3, 5, and 6. Screening estimates are not disclosure-grade
  • Market-based Scope 2: Evidence-gated — requires tenant-held contractual instruments (GreenPower evidence). Does not replace current location-based outputs

Data boundaries​

  • Physical risk: Based on observed BOM SILO historical data. CMIP6 scenario projections supported when ingested (ADR-091). Water stress from BOM water balance when ingested (ADR-095)
  • Entity resolution: Probabilistic matching via Splink. Low-confidence matches are flagged, not suppressed
  • Climate bands: Use a 1981–2010 baseline period. Drought modifiers escalate bands based on observed rainfall
  • Site-level data: Requires imported entity sites with geocoded locations

Disclosure boundaries​

  • Governance and Strategy: Placeholder-only sections — require entity-authored narrative
  • Scenario analysis: CMIP6 projection support implemented (ADR-091); populates when projection data is ingested
  • Transition risk: Config-driven sector + Safeguard + benchmark scoring (ADR-093)
  • Filing readiness: AASB S2 pillar-level progress tracking available (ADR-090) — not legal sufficiency
  • Framework coverage: Descriptive only — does not imply filing readiness or legal sufficiency
  • Targets: Storage and progress tracking are supported but intensity targets and Scope 3/1+2+3 targets are not progress-assessed in the first release

Rollup boundaries​

  • Default rollup (public): Company emissions from publicly available NGER data
  • Declared boundary (declared_boundary): Requires tenant-authored organisational boundary. Equity-share weighting applies to company emissions math only — site-level outputs are not ownership-weighted
  • Peer benchmarking: Sector benchmarking with percentile rank available (ADR-088); requires minimum 10 entities per ANZSIC division

Operational boundaries​

  • Batch limits: 50 ABNs per batch request, 200 factor lookups per batch
  • Disclosure generation: Async by default — returns job ID for polling
  • Artifact retention: Disclosure bundles are stored and versioned per company/tenant/pack scope
  • Amendment tracking: Factor set changes are versioned but existing disclosure artifacts are not retroactively regenerated