Struck
The sky keeps throwing matches. We count where they land.
Channel: Struck
Tagline: The sky keeps throwing matches. We count where they land.
Niche: Consumer-facing lightning intelligence — an automated, visual, opinionated site that turns satellite lightning detections, storm-event records, wildfire hotspots, and fatality data into county scorecards, seasonal risk maps, wildfire-ignition explainers, climate-shift analysis, and brutally useful articles for people who work, live, travel, or play under open sky.
Target audience: Outdoor workers, farmers, golfers, hikers, anglers, boaters, pilots, storm nerds, local reporters, wildfire watchers, homeowners in storm belts, and the small-but-lucrative class of people who buy weather stations, emergency gear, and insurance because they’ve learned the hard way that weather is not a vibe — it is a bill.
Why now: Three reasons. First, the open-data stack is suddenly good enough: NOAA’s GOES lightning mapper is live on AWS, NASA FIRMS gives fire hotspots, and NCEI keeps decades of lightning event records in bulk CSV. Second, public curiosity is real but underserved: people search for lightning maps every storm season, yet the market is dominated by dot-cloud visualizers with zero analysis. Third, the commercial tail is stronger than it looks: lightning sits at the intersection of wildfire, property damage, outdoor safety, insurance, climate anxiety, and shareable visual spectacle. This is exactly the kind of niche where useful obsession beats generic weather coverage.
Content Example:
Sample headline: Dry Lightning Is the Most Expensive Kind of Drama in the West
The public thinks lightning is a theater problem — bright sky, loud boom, maybe don’t stand under a tree. The data says it is a systems problem. When lightning arrives without meaningful rain, it stops being a moment and becomes an invoice: for forests, for utility crews, for insurers, for counties that suddenly discover their evacuation routes were designed by optimists. Dry-lightning storms don’t just start fires. They start clustered fires, scattered across terrain too awkward for fast response, at exactly the moment the air is hot, fuels are receptive, and everybody who was underprepared yesterday is now calling it an “unprecedented event.”
This is where Struck earns its keep. A normal lightning map tells you where the sky snapped. We tell you which strikes landed in drought-stressed fuel, how often that county has been hit in the last ten seasons, whether the area’s wildfire detections tend to appear within the next 24 to 72 hours, and whether local risk is drifting upward. The useful question is not “Did lightning happen?” Of course it did. The useful question is: “When lightning hits here, what tends to happen next?” That is the difference between weather content and weather intelligence.
Data Sources:
- NOAA GOES-16 / GOES-19 GLM — satellite lightning flash/group/event data from the Geostationary Lightning Mapper, openly available via AWS Open Data. This powers near-real-time strike density maps, flash energy estimates, storm cluster tracking, and regional seasonal flash counts.
- NCEI Storm Events Database — decades of U.S. lightning records with fatalities, injuries, damage, county/city information, and narrative text. This powers historical scorecards, “deadliest counties,” activity-based fatality patterns, and long-term regional comparisons.
- NASA FIRMS API — fire hotspot detections from MODIS/VIIRS for lightning-to-fire correlation and “possible ignition cluster” reporting.
- NWS Lightning Fatality tables — annual fatality tracking with age, sex, activity, and location. This powers safety journalism that is actually specific, instead of recycled “avoid open fields” boilerplate.
- NOAA SWDI / storm archives — additional severe-weather access formats for spatial overlays and archive work.
- Optional secondary inputs: U.S. Drought Monitor, HRRR/CAPE model fields, land-cover data for improved ignition-risk storytelling.
Automation Pipeline:
- Schedule:
- Every 30–60 minutes during storm season for fresh lightning summaries
- Nightly for county/corridor scorecards and wildfire-correlation updates
- Weekly for long-form “Storm Ledger” analysis pieces
- Monthly for climate/trend pages and rankings
- Collect:
- Pull latest GOES GLM files from AWS
- Fetch NASA FIRMS hotspot CSVs for relevant geographies
- Ingest NCEI monthly lightning-event CSVs and NWS fatality table updates
- Join with drought/land-cover/region metadata
- Process:
- Deduplicate flashes into meaningful storm clusters
- Compute rolling county, metro, and corridor lightning density scores
- Match likely lightning-fire sequences by space + time windows
- Extract narratives from Storm Events and summarize recurring patterns
- Generate editorial angles: “where lightning risk is rising,” “where the same outdoor activities keep getting people killed,” “which regions get spectacle vs. consequence”
- Generate:
- Lightning density heatmaps
- Dry-lightning risk maps
- Fatality activity charts
- Seasonal comparison charts by county/state/region
- AI-generated hero illustrations in a signature electric-noir style
- Auto-written article drafts with citations, then auto-graded for evidence density and rewritten for clarity/tone
- Publish:
- TypeScript static site build (Astro + D3 + MapLibre)
- Precompute JSON/GeoJSON data artifacts in GitHub Actions
- Deploy to GitHub Pages or Cloudflare Pages
- Optional Telegram bot / newsletter excerpt pipeline from the same content graph
Tech Stack:
- Static site: TypeScript + Astro
- Interactive graphics: D3.js, Observable Plot, MapLibre GL
- Data collection: Node.js scripts, AWS CLI/S3 sync for GOES, CSV/NetCDF parsers
- Storage: Flat-file JSON/GeoJSON in repo artifacts or lightweight object storage
- Image generation: AI illustrations for hero art; programmatic charts/maps for trust-building visuals
- CI/CD: GitHub Actions on cron + manual dispatch for major storm days
- Hosting: GitHub Pages initially; Cloudflare Pages if traffic grows
Monetization Model:
- Channel 1: donations/tips — weather nerds, storm chasers, and people in high-risk regions donate to tools that are more useful than generic media
- Channel 2: newsletter premium — paid weekly “storm risk memo” for photographers, outdoor professionals, wildfire watchers, and local journalists
- Channel 3: affiliate/sponsored — weather radios, lightning detectors, home surge protection, backup power, emergency kits, outdoor gear, weather stations
- Channel 4: B2B-lite sponsorship — insurer-adjacent newsletters, emergency-prep brands, weather-tech tools, wildfire readiness products
- Projected month-1 revenue: $150–$500
- Projected month-6 revenue: $1,500–$4,000 with search traction + storm-season virality
- Unit economics: Cheap. Core public datasets are free. Main costs are build time, occasional storage/bandwidth, and AI tokens for writing/imagery. This is a high-information, low-infrastructure niche.
- Email list value: Stronger than average because subscribers self-identify as risk-aware, gear-buying, locality-sensitive readers — exactly the people advertisers and premium products want.
Launch Complexity: 4/5 — medium. The data is available, but GLM handling and lightning-fire correlation take real work. Estimate: 4–6 focused days for a v1 with maps, scorecards, and one good weekly article format.
Content Quality Score: 5/5 — genuinely useful, visually distinctive, and unusually explainable from hard data.
Automation Score: 5/5 — the pipeline is cron-native and highly machine-friendly once built.
Revenue Potential: 4/5 — not a mass-market monster, but a very monetizable niche audience with clear sponsorship and premium-report angles.
Total: 18/20
Why This Will Work: The best automated channels do one thing better than the internet’s sludge pile: they answer a real question with evidence and style. Struck answers several. Where is lightning actually getting worse? Which places get spectacle versus actual damage? Which outdoor behaviors keep ending in funerals? Which counties should worry about dry lightning as a fire starter, not just a scary sky effect? This channel wins because it converts raw weather telemetry into consequence. That is catnip for SEO, for social sharing, for local-news pickup, and for niche donors who love finding a source that is both smarter and more alive than government PDFs.
There is also a strong personality layer here. Struck should sound like a storm chaser who learned statistics and got tired of weather coverage written by people who think “severe” is an adjective instead of a dataset. The site’s visual identity is electric-noir: black sky, white type, acid-yellow alerts, violet strike maps, animated branching motifs, recurring features like Flash Count, Dry Trouble, and The Sky Filed Another Complaint. That personality is what turns a useful niche tool into a supportable media property.
Risk & Mitigation:
- Risk: Real-time lightning maps already exist.
Mitigation: Do not compete on “look, a flashing dot.” Compete on analysis, local scorecards, wildfire correlation, and written intelligence. - Risk: Global real-time coverage is messy and fragmented.
Mitigation: Start with the U.S. + adjacent regions where NOAA/NWS/NASA data is strongest, then expand. - Risk: Severe-weather interest is seasonal.
Mitigation: Lean into off-season explainers, annual rankings, climate trend pages, and evergreen county profiles. - Risk: Over-automation can become weather slop.
Mitigation: Enforce article templates that require comparisons, consequences, cited sources, and at least one non-obvious insight before publication.