European aerodrome weather observability
See when the report arrived—and what changed.
Terminal Weather Services is developing an independent data-quality layer for METAR and SPECI distribution. The proposed evaluation design will preserve the message returned by each source, separate its timestamps, and turn each observed revision into a traceable station event.
- 90-day proposed evaluation
- 12 benchmark aerodromes
- Non-safety-critical
Working prototype
Public-feed monitor
This view queries all benchmark stations in one server-side request and refreshes no more than once per minute. It demonstrates ingestion, timestamp handling and raw-message presentation—not licensed-provider performance.
receiptTime − obsTime: the interval between the observation timestamp and receipt by AWC. It is not end-to-end provider latency.
| Station | Observation | AWC receipt | Receipt delay | Report |
|---|---|---|---|---|
| Requesting the current public-feed snapshot… | ||||
A snapshot alone cannot establish whether a scheduled report is missing or late. Station schedules and source-specific receipt paths are required for that classification.
What this proves
A functioning ingestion path
Combined station retrieval, source timestamp separation, TAC presentation, basic report flags, cache-aware refresh and a normalized browser view.
What it does not prove
Operational-feed performance
The public prototype does not represent an airport origin feed, guaranteed delivery, independent provider comparison or suitability for operational decisions.
Purpose-built scope
One observability layer, four defined jobs.
Receipt timing
Keep source and platform timestamps separate, then measure defined intervals by station and upstream path.
Message integrity
Retain each raw message observed by the platform, identify it by hash, and trace corrections and revisions.
Station events
Normalize new reports, corrections, NIL states and source changes into a consistent event envelope.
Distribution assurance
Evaluate completeness and consistency against station-specific schedules and agreed source behaviour.
Proposed evaluation architecture
Keep the evidence before shaping the data.
The source message and timing envelope enter the archive before normalization. That order makes revisions explainable and prevents decoded fields from replacing the source evidence.
- Source-specific adapters and receipt clocks
- Append-oriented raw-message archive
- Hashing, validation and revision linkage
- Normalized station-level event stream
Proposed provider pilot
A contained 90-day technical evaluation.
Small enough to configure deliberately; long enough to observe routine reports, SPECI events and corrections across different national paths.
-
Days 1–15
Connect and establish
Confirm licence boundaries, map source metadata, validate clocks and establish a completeness baseline.
-
Days 16–60
Observe and compare
Measure defined receipt intervals, catalogue revisions and investigate station- or path-specific exceptions.
-
Days 61–90
Review and decide
Produce a provider performance summary, verify permitted retention and agree whether a commercial phase is justified.
Benchmark coverage
Twelve aerodromes, one controlled station set.
Madrid is the priority path. Stockholm anchors the benchmark locally; the remaining stations provide a practical cross-section of distribution paths across Europe and Türkiye.
Evaluation controls
Licensed data stays inside the agreed boundary.
Controls will be finalized against the provider contract before licensed data is connected. The baseline is intentionally simple: restrict access, preserve an audit trail, minimize personal data, and delete on a defined schedule.
Access and transport
TLS in transit, environment-managed secrets, least-privilege access and MFA on administrative accounts.
Retention
Evaluation data retained for the evaluation term plus 30 days unless the contract requires less; backup deletion completes within a further 30 days.
Raw-data boundary
Raw licensed messages will not be displayed publicly, supplied to customers or redistributed without explicit written permission.
Purpose
Non-safety-critical data research, quality analysis and infrastructure evaluation. Not aircraft navigation, communication or control.
Provider discussion
Evaluate a source against a defined, limited scope.
We are looking for original TAC METAR and SPECI messages, unbatched or event-driven delivery, and source-receipt metadata where available. A modest budget is available for a paid evaluation, subject to scope and quotation.
- Coverage
- 12 benchmark aerodromes
- Term
- 90 days proposed
- Use
- Internal evaluation only