Weather Impact Intelligence

WEATHER IMPACT INTELLIGENCE DESTINATION-AWARE ROUTE CANVASS PLANNER v0.87

Area-level prospecting intelligence • not verified property damage

FIELD DATA STATUS • Validated affecting scoring: 0 • Observed awaiting validation: 0 • Test/demo + legacy excluded: 0 • Registry: v3

Targets near address

Automated opportunity discovery

Where should the team look first?

The discovery engine now radar-screens NWS warning/cell geography even when no SPC/NOAA point report has arrived yet. MRMS/MESH can therefore create the first observed hail anchor; warnings themselves still do not count as verified or localized damage evidence. Radar/intensity enrichment is then shared back to time-compatible territories before field deployment.

Loading 1d • run #— • starting…

Export Top Opportunities

Smart Radar: WII processes the first 0 highest-priority weather/radar seeds in parallel with territory enrichment and reuses historical MRMS results from persistent cache. Full 0-scan enrichment is optional.

Opportunity zones

…

0 underlying territories

Storm signals

…

distinct recent weather episodes

Radar screened

PENDING

priority radar scans • parallel with enrichment

Deployment-ready zones

…

Deploy + Strong Candidate

Weather window

24h

hail + wind

Weather Data Reliability

Source-by-source ingestion health and recovery

A failed source no longer silently becomes “no storm.” SPC reports, NWS alerts and Census geography are diagnosed independently, and usable sources continue feeding discovery whenever possible.

LOADING WEATHER SOURCES

SPC / NOAA reports

PENDING

…

qualifying Cincinnati/Dayton reports

0 report feeds with matches • 0 empty/404 • 0 failed

NWS alerts

PENDING

…

relevant watches/warnings/statements

Census geography

PENDING

…

micro-territory block groups loaded

0 ZIP/ZCTA areas available

Recovery mode

Multi-source discovery

SPC evidence can create footprints/cells without NWS. NWS warning polygons can create candidates without SPC. Geography failures are surfaced separately rather than misreported as “no storm.”

Source request diagnostics

SPC daily feed attempts

NWS state requests

Used window: — → —

Window-aware storm episode acquisition

0 storm episodes identified across the selected window

WII now constructs independent storm episodes across the entire selected window before allocating the MRMS budget. Each selected episode receives temporal coverage based on severity, Cincinnati / Dayton market intersection, residential relevance, uniqueness and recency; one storm can no longer consume the whole radar budget.

0/0 selected episodes complete • 0/0 scans

QUALIFYING EPISODES

SELECTED FOR MRMS

0/0

EPISODES COMPLETE

0/0

RADAR SCANS

Detected Storm Cells

Damage-signal discovery before territory ranking

Cells are built from geographically and temporally related SPC/NOAA reports plus NWS warnings that contain meaningful wind/hail/tornado signals. These cells are screening seeds—not verified property damage.

0 cells detected

No qualifying damage-signal cells were detected in the selected weather window. Territory rankings should therefore be treated as broad storm-envelope screening only.

Storm Track Intelligence

Warning envelope → time-ordered report track → radar/core evidence

NWS warning magnitude alone cannot create Moderate/Strong exposure.

Warning envelope

candidate territories

Report ≤5 mi

near-core corroboration

Report ≤3 mi

core proximity

Core report

report within 3 mi

High core confidence

strong direct evidence

Radar localized

shared MRMS episode footprint

Track core ≤2 mi

Near core 2–5 mi

Fringe 5–8 mi

Envelope only

Observed-Evidence Damage Cores

Observed weather anchors before neighborhood ranking

A physical core can now be seeded only by observed localized evidence: MRMS/MESH hail or a nearby direct SPC/NOAA damaging-weather report. Inferred tracks, generated cells and warning polygons may shape or corroborate a core, but cannot create one. Every supporting territory remains bounded to the observed anchor. Property size and likely-door count do not create or strengthen weather confidence.

0 independent cores from 0 localized territories

Independent cores

physical opportunities

High-confidence cores

confidence ≥72

Wind / mixed cores

wind-dominant opportunities

Hail / mixed cores

hail-dominant opportunities

No localized physical damage cores were formed in this weather window. Broad storm-envelope territories remain available for diagnostic review but are not promoted as independent damage opportunities.

Storm Footprint Funnel

From broad storm association to localized field opportunity

Evidence confidence, not territory size or broad warning overlap, now gates deployment priority.

Underlying territories

storm-linked + 25+ doors

Opportunity Zones

practical canvass clusters

Deployment-ready zones

Deploy + Strong Candidate

Strong / Verified evidence

localized confidence

Moderate / Preliminary

validate before dispatch

Radar screened

0/0

priority MESH checks

Radar Diagnostics

MRMS/MESH footprint localization

Every radar seed is carried from archive lookup through decoded MESH grid cells, connected hail-footprint clusters, and actual territory intersection. A radar hit elsewhere in the 55-mile crop no longer looks like a localized target. Exposure and deployment thresholds are unchanged.

0/0 completed

Event Intelligence v2

Multi-source storm awareness

Authoritative weather remains the primary evidence layer. Ground/news/social sources are treated as corroboration and will not be allowed to masquerade as verified meteorological observations. NWS warning polygons now expand storm-path candidate discovery; MRMS remains the refinement layer before deployment.

NWS alerts degraded

SPC / NOAA reports

LIVE

primary hail/wind report evidence

NWS watches / warnings

DEGRADED

0 warning-level alerts • 0 with parsed wind/hail intensity

Local news corroboration

DEGRADED

GDELT local-damage/news signals

MRMS / MESH

AUTO

0/0

radar enrichment of discovery seeds

Planned corroboration sources

Integration-ready

Official API credentials/credits required before production ingestion.

Bluesky

Research candidate

Public AppView/search architecture is promising for low-cost public corroboration.

Commercial approval needed

Do not rely on the standard Data API for a commercial product without appropriate agreement.

Mastodon

Research candidate

Status search availability varies by instance and authentication/search backend.

Power outages

Enterprise source

Potential high-value ground-impact corroboration; commercial API access is separate.

Discovery logic: v0.84 preserves synchronized discovery, rep-session state, and travel economics, then unifies Google address lookup behind one resilient adapter. It tries Places API (New), falls back to Places API (Legacy), and surfaces Google authorization/configuration errors directly in the form instead of silently hiding them. This remains area-level prospecting intelligence—not verified property damage.

Admin discovery diagnostics

LOADED MICRO

AFTER EVENT FILTER

DIRECT EVIDENCE

STORM LINKED

DOOR ESTIMATE

MISSING DOORS

CANDIDATES

RADAR QUEUE

Micro targets are already generated by the live-target service from storm-report screening footprints. A direct point-report inside every Census block group is therefore not required for initial discovery. Radar remains a second-stage enrichment step.

Loading discovery candidates…

Live NOAA/Census territory data is still loading.