Interactive asteroid impact-effects simulator using NASA/JPL close-approach data
Asteroid Map lets you drop any space rock on any city and explore a first-order impact-effects scenario. Pick a location on the interactive globe, choose from real near-Earth close-approach records pulled live from NASA/JPL's database (or famous historical impactors like Chicxulub), and see modeled effect zones overlaid on the map — crater, fireball, blast wave, window damage, and thermal radiation zones — along with broad uniform-density population exposure ranges and energy release.
The physics use first-order scaling from Collins et al. 2005 "Earth Impact Effects Program". These are educational estimates, not emergency forecasts or precise survival predictions.
- Real asteroid data — live close-approach objects from NASA/JPL's SBDB Close-Approach Data API, filtered by size and approach date; some diameters are estimated from brightness
- Famous impactors — preloaded with Chicxulub, Apophis, Tunguska, Chelyabinsk, and more
- Five effect zones visualised on the map:
- Crater (excavated-ground radius)
- Fireball (extreme thermal exposure)
- Blast wave (about 4 psi structural-damage screening)
- Window damage (low overpressure — broken-glass injury risk)
- Thermal radiation (3rd-degree burns)
- Exposure, casualty, and uninjured-survivor ranges based on explicit uniform-density scenarios, a custom people/km² value, or optional WorldPop ring-population estimates, broadened by mapped-radius uncertainty
- WorldPop exposure assist — optional lookup estimates residential population by mapped effect ring using WorldPop 100 m gridded population data
- Scientific readout — impact energy in megatons, seismic energy equivalent, modeled event type, approximate global frequency, and uncertainty notes
- Shareable URLs — simulation inputs are encoded in the URL so you can share the same scenario
- Mobile-friendly — step-through drawer UI on small screens
- Scenario controls — choose the impactor, target location, impact angle, target material, and uniform population-density scenario
- Speed handling — live CAD v-infinity values are converted to modeled impact-entry speeds with Earth-gravity focusing; close-approach relative speed is only a fallback
| Layer | Technology |
|---|---|
| Framework | Next.js 15 (App Router, server + client components) |
| Language | TypeScript 5 (strict mode) |
| Styling | Tailwind CSS 4 |
| Map | MapLibre GL + React Map GL |
| State | Zustand |
| Geocoding | OpenStreetMap Nominatim (proxied via API routes) |
| Asteroid data | NASA/JPL SBDB Close-Approach Data API |
| Icons | Lucide React |
- Node.js 20.11 or later
- npm (included with Node.js)
# 1. Clone the repo
git clone https://github.com/ZipLyne-Agency/Asteroid-Map.git
cd Asteroid-Map
# 2. Install dependencies
npm install
# 3. (Optional) Set up environment variables
cp .env.example .env.localAll optional — the app runs with none set. Create a .env.local in the project root to override:
# Optional — identifies your instance to the OSM Nominatim geocoding service
# Replace with your contact info as required by the Nominatim usage policy
NOMINATIM_USER_AGENT=AsteroidMap/1.0 (your@email.com)
# Optional — override the canonical site URL (defaults to http://localhost:3000 in dev)
# NEXT_PUBLIC_SITE_URL=https://asteroidmap.comnpm run dev # Start dev server at http://localhost:3000
npm run build # Production build
npm start # Serve the production build
npm run lint # Lint with ESLint
npm run typecheck # TypeScript checks
npm test # All regression tests
npm run check # Lint, typecheck, tests, and production build├── app/
│ ├── api/
│ │ ├── geocode/reverse/ # Reverse geocoding (lat/lng → place name)
│ │ ├── geocode/search/ # Forward geocoding (place name → coords)
│ │ └── neo/approaches/ # NASA near-Earth object data
│ ├── layout.tsx # Root layout, metadata, fonts, JSON-LD
│ └── page.tsx # Main page
│
├── components/
│ ├── DirectoryMap.tsx # Interactive map with impact zone overlays
│ └── JsonLd.tsx # Structured metadata
│
└── lib/
├── physics.ts # Impact calculations (airburst, crater, blast, thermal, casualty ranges)
├── store.ts # Zustand global state
├── known-asteroids.ts # Preloaded asteroid database
└── major-cities.ts # Major city coordinates for quick selection
Impact effects are calculated with first-order scaling from the Earth Impact Effects Program (Collins et al. 2005). Key formulas:
- Crater radius — scales with impactor kinetic energy and target density
- Fireball radius — extreme thermal exposure screening from energy-yield scaling
- Blast overpressure — about 4 psi structural-damage screening from yield scaling
- Window damage — low-overpressure broken-glass injury screening
- Thermal radiation — fluence threshold for 3rd-degree burns
- Exposure, casualty, and survivor range — modeled ring populations × broad heuristic vulnerability rates, with mapped-radius uncertainty included in the low/high range
- Recurrence interval — approximate global frequency from NASA hazard-scale size anchors
Calibration references include the Collins/Melosh/Marcus Earth Impact Effects Program, NASA's Chelyabinsk summary (about 440 kt TNT, 14 miles burst height, windows blown out over about 200 square miles, and over 1,600 mostly broken-glass injuries), NASA/JPL CAD documentation for live close-approach fields, and Glasstone/Dolan-style air-blast scaling for overpressure screening. The app keeps these as broad calibration checks rather than exact guarantees.
This app cannot make exact survival predictions. Real consequences depend on fragment strength, breakup altitude, terrain, weather and visibility, sheltering, building stock, time of day, emergency response, actual target geology/water depth, and actual gridded population distribution. Fatality and injury outputs are illustrative screening scenarios that apply broad heuristic vulnerability rates to effect-zone populations; they are not validated casualty forecasts. The casualty engine can accept population totals per effect zone, which is the correct shape for raster-derived exposure. The current UI uses a user-selected uniform density scenario, a custom people/km² value, or an optional WorldPop lookup that estimates population separately for the mapped effect rings; it does not yet use LandScan-style day/night population rasters or building-level vulnerability. Low/high ranges expand using the same mapped-radius uncertainty shown in the report. For best exposure inputs, use a local census product or gridded population dataset such as WorldPop 100 m, Kontur Population 400 m H3, or NASA SEDAC GPWv4 1 km. For NASA/JPL objects, casualty ranges are widened with the measured 1-sigma diameter uncertainty when supplied, or with the estimated diameter range when CAD lacks a measured diameter. Airburst handling is approximate and suppresses crater/seismic effects when the object disrupts above the surface. Water targets only adjust crater scaling; tsunami, water depth, and coastal run-up are not modeled. NASA/JPL close-approach records are real flybys, not predicted impacts; when CAD lacks a measured diameter, the app estimates a diameter range from absolute magnitude and an assumed albedo span, then uses the midpoint estimate for the scenario. Diameter-range uncertainty is reflected in the displayed mass and energy ranges because both scale with diameter cubed. For CAD speeds, the modeled impact-entry speed uses v-infinity with Earth-gravity focusing when JPL supplies it; close-approach relative speed is only used as a fallback. When a curated object has a published bulk density, the app uses that value for mass and energy; otherwise it uses the selected material-density bucket.
Contributions are welcome! Here's how to get involved:
- Fork the repository and create your branch from
main - Make your changes — run
npm run lintandnpm run buildbefore submitting - Open a pull request with a clear description of what you changed and why
- Adding more historical impactors to
lib/known-asteroids.ts - Improving mobile UX
- Internationalisation (i18n)
- Accessibility improvements
- Unit tests for
lib/physics.ts
Open an issue and include:
- Steps to reproduce
- Expected vs. actual behaviour
- Browser and OS
The app uses the standard Next.js Node deployment. Any platform with Next.js support works (Vercel, Railway, Fly.io, etc.).
npm run build
npm startFor containerised deployments, add Next.js standalone output and copy public and .next/static into the runtime image as described in the Next.js deployment documentation.
Set NEXT_PUBLIC_SITE_URL to the public HTTPS origin when deploying somewhere other than https://asteroidmap.com. No database, account, payment provider, or private API key is required.
- NASA/JPL close-approach records are retrieved server-side from the public SBDB CAD API.
- Place searches and reverse geocoding are proxied to OpenStreetMap Nominatim.
- WorldPop population lookup is optional and sends the selected coordinates and modeled radii to the WorldPop API through the server.
- On first visit the app requests browser geolocation to center the map. If that is denied, it falls back to a curated city for the browser timezone (New York if unknown).
- The browser loads map styles and tiles from OpenFreeMap. OpenFreeMap therefore receives ordinary web-request metadata such as IP address and user agent.
- Simulation choices are kept in the page URL so a scenario can be bookmarked or shared. Do not share a precise location if you consider it sensitive.
- The application itself has no accounts, analytics, advertising, or cookies.
See PRIVACY.md for details and THIRD_PARTY_NOTICES.md for data/service attribution.
Please report vulnerabilities privately as described in SECURITY.md. Do not put sensitive exploit details in a public issue.
MIT — see LICENSE for details.
Built with curiosity about things falling from space. Visit asteroidmap.com.