Weather and Disaster LabLive risk checks from NOAA, NWS, and public hazard feeds

Your weather & hazard view

Weather Lab

See live source posture, short-term risk checks, medium-term hazard watches, and long-term climate/disaster signal hypotheses.

Start here

Where should we check weather risk?

Allow location once for a personal hazard snapshot. Coordinates stay in this browser tab only — never written to KELLY ground truth.

location off

Opt-in only. Not emergency alerts — follow NWS and local authorities.

New here? Start in 3 steps

Check weather risk near you

Like Apple Weather and Weather.gov, we lead with your area first — then show proof behind every read.

Step 1

Share your location

Use GPS or search for your city above. Coordinates stay in this browser tab only.

Step 2

Check your hazard snapshot

See risk level, time window, and active triggers in the live hazard card below.

Step 3

Ask or pick a preset

Use a quick question for 24-hour, storm, or long-range risk — every answer links to proof.

Not emergency advice. Follow NWS, USGS, and local authorities for warnings.

Your local snapshot

Live hazard watch near you

Risk level, time window, and active triggers from official feeds. Education only — follow NWS and local authorities for warnings.

Composer: · Pin ready:

Feed-backed nodes:

Content hash:

checking
checking
checking
checking
See official alerts

Intensity & window

Waiting for first card refresh.

Time window:

Location:

Feed honesty:

Updated from NOAA & public sources

Ask KELLY about this evidence

Behind the snapshot

Extra proof when available

connecting

Checking for additional proof items…

No live proof items right now — that's normal. Your hazard card above uses the same official feeds.

proof routeswaiting for first check
Advanced: audit trail, routes & technical proof
What KELLY is doing

This window shows what KELLY is doing in Weather Lab: what it is reading, what it thinks might matter, what proof exists, and what is still waiting.

How to audit it

Read the answer in this lab first. Open the proof shelf or JSON packet for the evidence behind it. Check the route status and last-checked timestamp before trusting freshness. Look for the mind-changer: the fact that would make KELLY revise the answer. If evidence is waiting, treat the output as a question or watch item, not a settled answer.

Evidence status

Live evidence appears only when KELLY exposes proof-linked sector output. Empty states are intentional waiting states, not filler.

What changes KELLY's mind

New approved evidence, a falsifier, settlement, replay drift, calibration movement, or memory-lineage revision.

What this window cannot do

This lab communicates with KELLY as an inspection window. It cannot write ground truth, memory, GraphState, provider state, or reality receipts.

Processing flow

How KELLY tracks weather risk with proof

01Read approved weather, water, fire, geography, and hazard sources with source freshness preserved.read sources
02Separate observation, region, hazard type, forecast horizon, uncertainty, and official alert authority.find your area
03Emit hazard watch cards, source freshness receipts, forecast deltas, and closure windows.forecast risk
04Route settled official outcomes into replay, calibration, and memory-lineage eligibility without becoming emergency authority.check reality

Sources

Official sources we watch

  • NOAA and National Weather Service
  • NOAA water and hydrology APIs
  • NASA hazard and earth observation feeds
  • GeoNames location context
  • Open-Meteo and public weather feeds

Proof standard

How users can tell

  • Private user location can guide a personal view without writing Prime ground truth.
  • Prime KELLY stays worldwide and source-governed.
  • Forecasts must name source freshness and uncertainty.
  • Reality closure checks official outcomes, not user preference.

Outputs

What you can see here

  • hazard_forecast_card
  • hazard_watch_cards
  • source_freshness_receipts
  • regional_risk_context
  • forecast_delta_receipts
  • official_outcome_closure_candidates

Storage posture

Postgres: central KELLY spine

Use canonical KELLY Neon/Postgres for hazard_observations, region_context, forecast_deltas, source_freshness, closure_windows, settlement, and calibration state.

Object storage

Immutable evidence shelf

Use R2/object storage for immutable forecast packets, alert snapshots, map/source bundles, and replay receipts.

Volume

No ground-truth volume

Optional local volume only for rebuildable map tile, geocoder, and model cache. It must never become canonical hazard memory.

Technical objects

Primary objects

  • region
  • hazard_type
  • source_freshness
  • closure_window

Proof receipts

Black-box recorder

  • official_alert_receipt
  • observation_receipt
  • forecast_delta
  • settlement_receipt

Authority

Write boundary

  • KELLY is the only write authority for durable learning.
  • Labs are read-only processors and inspection windows.
  • Glass Lab remains the traceable evidence console.
This is not emergency, evacuation, medical, or safety advice. Always follow official alerts and local authorities. Public users can ask, inspect, export local viewport data where supported, and follow proof links. They cannot edit KELLY Prime ground truth from this app.