Case overview
Administrator NOTAM launch watch profile
Suggests case matches for public-warning leads. Discovery does not require this profile or an existing case. No warning becomes evidence until accepted. Include downrange areas, not just the launch site.
Public launch reference library
Public catalog history provides a transparent comparison for the raw NOTAM geometry. Your cases are not used by the estimator.
Open monthly GCAT checks ↗ · Download the gcat-review artifact and select candidate-reference-pack.json. Loading a file only previews it.
Review candidate catalog
Search public launch records and sources
Develop the case by phase
Add the evidence available at each phase. SHERLOC analysis is optional.Dated assessments
Save the current cutoff, assessment, inclination and evidence without locking the phase.
Evidence
Use one evidence item for each NOTAM, thread, source, or distinct piece of launch information. Images and files remain associated with that item.
NOTAM evidence
Add each NOTAM as a separate item.
Forum / source material
Add each thread, post, article, or source as a separate item.
Additional Launch Evidence
Add each distinct piece of orbit, vehicle, timing, payload, or analyst information separately.
Expected inclination
Estimate inclination from the NOTAM coordinates, or replace it with a documented estimate from another source.
Show calculation, formulas, and sources
No calculation recorded.
Method: the app derives a raw great-circle estimate from the launch site and NOTAM closure geometry. It then compares that result with earlier public catalog launches from the same site, prioritizing matching vehicles, pads, orbital regimes, and nearby inclinations. The raw and public-history results remain visible separately.
SHERLOC and analyst findings
Notes and attachments
When evaluation is enabled, preserve exactly what SHERLOC returned. Paste text or attach images, CSV, TSV, Excel, or PDF files.
Earlier response status
Accepted attachments: images, CSV, TSV (.tsv), Excel (.xls/.xlsx), and PDF. Files are preserved as submitted; spreadsheet and PDF contents are not automatically interpreted.
Earlier solution entries
Legacy solution entries
Earlier solution entries remain preserved. Enter the results for grading under SHERLOC prediction. These legacy entries do not enter the new grades automatically.
Enter only what SHERLOC predicted. Count-only predictions can leave payload names blank. A sole solution becomes the primary prediction automatically.
Phase summary
Record the verified outcome
Keep the verified outcome separate from the evidence used to develop the case.Actual controlled payloads
Enter one physical payload per row, or use one family name with a quantity when individual identifiers are unavailable. Accepted aliases are used only for identity matching.
Confirm only when your outcome evidence supports the identities. WATSON uses these identities to grade predictions.
Payload quantity (optional)
SHERLOC Metrics
Performance across the cases you can access.Performance by category
Accuracy = Correct ÷ assessable predictions. Partial results receive no automatic half-credit. Coverage shows assessed versus eligible recorded predictions. Inclusion rate counts Correct plus Partially Correct results. Quantities are descriptive only.
Performance over time
What remains unassessable or excluded
Cases in this view
Cases without SHERLOC evaluation remain listed. Open a case to see its evidence and comparisons. Archived rows open the case browser.
Default view: Pre-launch, active cases. Phases describe where predictions were recorded; predictions remain editable and are not proof of pre-outcome timing. Identical carried predictions count once when pooling phases. Retrospective interpretations are excluded from statistics unless selected.
Review and export
Review the case record and, when enabled, SHERLOC accuracy results.Export
Use Summary PDF for the case details, predictions, outcome and debrief. Detailed PDF includes the supporting evidence.
| Phase | Solutions | Count correct? | Complete package? | Shortlist score | Most likely |
|---|
Case chronology
Browse milestones by date. Expand evidence groups or an event for sources, notes and editing. Undated items and historical checkpoints are kept separately.
Add an analyst event
Debrief
ADMINISTRATOR TOOL
Review Queue
Candidate information stays outside cases until you accept it into a specific editable phase.Automatic NAVWARN collection
The Cloudflare Worker checks public maritime warnings every four hours. Explicit spaceflight warnings enter Pending; ambiguous rocket activity enters Possible launch. Routine navigation and clear missile/gunnery notices are excluded. Manual upload uses the same filter.
FAA launch-region collection
Automatically starts every day at midnight Colorado time (America/Denver), including daylight-saving changes. You can also start a run manually. Each run checks 37 flight-information regions (the original 32 plus five historical downrange queries) across China, Russia, Kazakhstan/Baikonur and selected downrange areas. You do not need to look up airport codes. Explicit spaceflight notices enter Pending review. Ambiguous rocket warnings and timed historical corridor matches enter Possible launch for separate review.
A full run takes about 110 minutes or longer because FAA requests must be spaced out. Collection continues when you close this page. Reopen Review Queue to see progress.
Regional results
View the 37-region watch map ↗
Mapped FAA review leads
Select a mapped notice to see its identifier and published timeframe.
Source polygons, lines and points only. A point is a published location, not an inferred restricted-area radius. This map follows the selected review view. Unrelated notices are excluded; possible matches are not confirmed launches.
FIR queries retrieve notices associated with regional codes. They do not retrieve every airport notice inside a country, and this watch list does not guarantee all launch corridors. A region does not identify a launch's country.
Optional: single-location diagnostic
Use these controls for a specific location or connection test. They share the regional run's FAA quota.
Coverage is limited to the chosen location; a FIR code may differ from an airport code. Accepted or rejected notices remain in history. Changed source content becomes a separate review lead. Published cancellations and replacements are linked when their reference matches a saved lead. Absence from a snapshot does not imply cancellation. Production tests and collection share a three-minute request limit.
Live CelesTrak SATCAT check
Administrator-triggered, one launch at a time. Confirmed U.S. launch-site records are excluded. The current catalog record is placed in Review Queue only when its object identity has not previously been reviewed or queued; it never becomes case evidence automatically.
Import a public maritime warning
Use a freely available TXT warning export or paste an openly accessible warning. Watson keeps only launch/re-entry-related matches and creates review leads; it does not fetch paid or sign-in-only sources.
Add a manual candidate
Add a lead manually when needed. Public maritime warnings can also arrive from the scheduled collector.
Add a review lead manually
Candidate decisions
Accepting a potential launch NOTAM adds it to the case’s NOTAM evidence; other candidates go to Additional Launch Evidence. Deferred and rejected items remain auditable, but are not case evidence.
Accepted and rejected leads move to Decision history automatically. Deferred leads stay pending. History retains the decision and source trail.
INDEPENDENT TOOL
Inclination Workbench
Evaluate NOTAM closure geometry without creating or changing a case.Launch context
Use the best site and cutoff information currently available.
NOTAM closures
Add each NOTAM separately. Text, source information, and attachments remain associated.
Inclination assessment
The workbench uses the same estimator and public catalog as every case phase.
Show calculation, formulas, comparison launches, and sources
No calculation recorded.
The raw NOTAM estimate, public-history comparison, combined result, manual override, calculation version, and source records remain visible separately.
Use the assessment
Export it as a standalone calculation record or copy it into an unfrozen case phase.