Global elevation data platform — query, analyze, and visualize terrain anywhere on Earth.
Live site: openzenith.org · Map · Globe · Explore · WASM Demo
API base: openzenith.cyopsys.com
OpenZenith is a geospatial SDK + REST API for working with global elevation data. It provides a unified Python interface for querying point elevations, loading terrain grids, and running terrain analysis (slope, aspect, hillshade, viewshed, hydrology) — with full offline support once local data is cached.
The key differentiator is low-latency local processing: no API calls needed for elevation queries when using local .merged tiles, and no complex DEM data setup required. Scientists, GIS engineers, and app developers get elevation data without managing SRTM/GEBCO tile downloads, coordinate conversions, or hydrology preprocessing.
Works entirely offline after installing the Python SDK and optional local data. Rust/WASM compute kernels handle CPU-intensive terrain analysis (D8 flow, viewshed, OZT2 decode) — running in-browser via WASM or as a local subprocess.
- Zero-config: pip install and query elevation in 2 lines of Python
- Offline-first: Local SRTM .merged tiles — no network required for elevation queries
- Low-latency: Rust/WASM compute kernels for D8 flow, viewshed, OZT2 decode — runs in-browser or subprocess
- Complete terrain analysis: slope, aspect, hillshade, viewshed, TPI, roughness, curvature, watersheds, stream extraction, downstream tracing
- Production-ready: Fully typed (py.typed), 1,650 pytest tests (99% coverage gate; last measured 99.11%), clippy-clean Rust, typed TypeScript API
git clone https://github.com/aliasfoxkde/OpenZenith && cd OpenZenith
pip install . # Core SDK
pip install .[all] # All extras: compression, download,
# analysis (aiohttp/rasterio/numba), viz, dev
pip install .[viz] # + matplotlib plotting helpersNote: the SDK is not yet on PyPI (publishing is maintainer-gated — see
docs/PUBLISHING.md), so the install is from source. Once the PyPI upload lands, these becomepip install openzenith[…].
import openzenith as oz
# Single point — uses HuggingFace tiles if no local cache
elev = oz.get_elevation(40.7128, -74.0060) # NYC → 10m
print(f"Elevation: {elev}m")
# Batch — parallel fetch with ThreadPoolExecutor
results = oz.get_elevation_batch([(40.7, -74.0), (48.8, 2.3), (35.7, 139.7)])from openzenith import load_elevation_grid, slope, aspect, hillshade
from openzenith.terrain import viewshed
# Load a 10km × 10km region around a point
grid = load_elevation_grid(lat=36.0, lon=-118.0, zoom=12, radius=200)
dem = grid['grid']
slope_map = slope(dem) # Degrees slope (Horn's method)
aspect_map = aspect(dem) # Compass degrees (0-360)
hillshade_map = hillshade(dem) # 0-255 analytical hillshade
# Viewshed from an observer point
visible = viewshed(dem, observer_row=100, observer_col=100, radius=50)from openzenith.hydrology import (
d8_flow_direction, flow_accumulation, extract_streams,
delineate_watershed, trace_downstream
)
flow_dir = d8_flow_direction(dem)
accum = flow_accumulation(flow_dir)
streams = extract_streams(accum, threshold=100)
# Trace downstream from a point
trace = trace_downstream(36.0, -118.0)
print(f"Distance to ocean: {trace['total_distance']:.1f} km")
# Delineate entire watershed
ws = delineate_watershed(36.0, -118.0)
print(f"Watershed area: {ws['area_km2']:.1f} km²")# Download SRTM tiles to ~/.cache/openzenith-dem/
openzenith download --region north-america --zoom-levels 0-10
# Or use HuggingFace directly (no local storage)
export HF_TOKEN=your_token_here| Module | Functions | Notes |
|---|---|---|
| elevation | get_elevation, get_elevation_batch, load_elevation_grid | Parallel tile fetching |
| terrain | slope, aspect, hillshade, viewshed, tpi, roughness, curvature, tri | Vectorized NumPy |
| hydrology | d8_flow_direction, flow_accumulation, extract_streams, delineate_watershed, trace_downstream, twi, fill_depressions | D8 algorithm |
| tracing | trace_downstream | Adaptive tile loading |
| fuse | FusedDEM | SRTM + GEBCO seamless |
| viz | plot_*, terrain_to_3d_mesh, terrain_to_glb, terrain_to_png | Matplotlib + 3D export |
| geotiff | export_geotiff, export_cog | GeoTIFF/COG export |
| merged | read_elevation_from_merged, MergedFile | OZCHNK01 local dataset access |
| async_client | ElevationClient, ElevationBatchProcessor | aiohttp batch concurrency |
| tile_format_v2 | encode, decode | OZT2 tiles (gradient prediction + adaptive quantization; Brotli/Zstd/zlib per tile) |
| converter | convert_tile, convert_directory | SRTM .merged → OZT2 pipeline |
| backends | OZT2Backend, OZT2HFBackend, OZT2R2Backend | Local files, HuggingFace, S3-compatible (the R2 class targets any S3-compatible store — the platform's own R2 bucket was decommissioned 2026-09-27) |
openzenith query --lat 40.7 --lon -74.0
openzenith trace --lat 36.0 --lon -118.0
openzenith watershed --lat 36.0 --lon -118.0
openzenith slope --lat 36.0 --lon -118.0 --radius 5km
openzenith hillshade --lat 36.0 --lon -118.0 --radius 5km
openzenith viewshed --lat 36.0 --lon -118.0 --height 10
openzenith fill-depressions --lat 36.0 --lon -118.0 --radius 10km
openzenith flow-accum --lat 36.0 --lon -118.0 --radius 10km
openzenith streams --lat 36.0 --lon -118.0 --threshold 100
openzenith info
openzenith validate
Base URL: https://openzenith.cyopsys.com/api/ — 81 API routes under api/src/app/api/.
GET /elevation?lat=40.7&lon=-74.0
POST /elevation/batch (up to 2000 points)
GET /slope?lat=40.7&lon=-74.0&radius=50
POST /watershed {"lat": 36.0, "lon": -118.0}
POST /trace {"lat": 36.0, "lon": -118.0}
GET /tile/{z}/{x}/{y} (terrain tiles)
Terrarium PNG elevation tiles are also served as an OGC WMTS 1.0.0 service, validated with GDAL as an independent client:
GET /tiles/WMTSCapabilities.xml (capabilities document)
GET /tiles/{tileMatrixSetId} (OGC API - Tiles metadata)
GET /dem-tile/{z}/{x}/{y} (WebMercatorQuad tiles)
GET /tiles/WorldCRS84Quad/{z}/{row}/{col} (true EPSG:4326 tiles)
Two tile matrix sets are advertised on separate layers:
WebMercatorQuad (EPSG:3857) and WorldCRS84Quad (OGC 17-083r2, true
lat/lon assembly — one layer per set, GDAL/QGIS-compatible).
Full API docs: https://openzenith.cyopsys.com/api/openapi.json
Python SDK (local compute) ←→ REST API (cloud, 81 API routes)
↓
Cloudflare Pages
↓
HuggingFace (origin) + edge Cache API
| SDK | Python 3.10+, NumPy, Rust (WASM + CLI) |
| API | Next.js 15, TypeScript, Cloudflare Edge |
| Data | SRTM 30m (HuggingFace), GEBCO 2025 |
| Tests | 1,650 pytest @ 99% gate (Python), 247 cargo test across feature sets (Rust core), 1,709 vitest across 115 files (TypeScript) |
Data Layers (62 mountable data layers, 62 curated in api/src/lib/layers/registry.ts — 31 highlights below)
| Layer | Source | Details |
|---|---|---|
| Earthquakes | USGS | Global seismic, magnitude + depth, timeline playback |
| Weather Radar | RainViewer | NEXRAD mosaic, animated precipitation |
| Weather Warnings | NOAA NWS | Watches, warnings, advisories (US) |
| Hurricane Tracks | NOAA NHC | Active cyclones with track animation |
| Wildfires | NASA FIRMS VIIRS | 3,000+ fire detections/day |
| Air Quality | Open-Meteo | AQI, PM2.5, PM10, NO₂, O₃ |
| Lightning | Blitzortung | Real-time global strikes (WebSocket) |
| SIGMETs/AIRMETs | NOAA AWC | Aviation weather hazards |
| Volcano Alerts | USGS | Real-time alert status |
| Disaster Alerts | GDACS | Global disaster aggregation |
| Flood Extent | Copernicus EMS | Flood monitoring |
| Sea Ice | NSIDC/OSI SAF | Polar ice coverage |
| Burn Scars | NASA FIRMS | Active fire perimeters |
| Space Weather | NOAA SWPC | Aurora forecast, Kp index |
| Layer | Source | Details |
|---|---|---|
| Flights (ADS-B) | OpenSky Network | 10,800+ live aircraft positions |
| Military ADS-B | ADSB Exchange | Military transponder data (requires key) |
| Vessels (AIS) | AISstream | Ship tracking (WebSocket) |
| Layer | Source | Details |
|---|---|---|
| NLNOG Nodes | NLNOG Ring | 750+ network measurement nodes |
| Buildings | Overture Maps | Global building footprints |
| Satellites | Celestrak | 1,500+ orbital positions, notable labels |
| Population Density | JRC GHSL | 100m resolution global |
| Land Cover | ESA WorldCover | 44-class land use |
| Waterways | HydroSHEDS/OSM | Rivers, lakes, water features |
| Layer | Source | Details |
|---|---|---|
| Hillshade | SRTM 30m | Terrain shading |
| Elevation Color | SRTM 30m | Hypsometric color ramp |
| Data Accuracy | Multi-source | Resolution heatmap (default on) |
| Topo Contours | SRTM 30m | 100m/500m contour intervals |
| Bathymetry | GEBCO 2025 | Ocean depth shading |
| Satellite Imagery | NASA GIBS | MODIS Terra true color |
| Night Lights | NASA VIIRS | Black Marble city lights |
| Marine Weather | Open-Meteo | Wave height, SST, wind |
- 10 basemaps — Dark, Dark+, Dark (no labels), Streets, Light, Positron, OpenStreetMap, Satellite, Topographic, Terrain
- Opacity sliders on all raster layers
- DD/DMS coordinate toggle in position panel
- Elevation profiling — SVG sparkline on distance measurement
- Annotations — draw points, lines, polygons with naming
- Bookmarks — save/restore viewport + layers to localStorage
- Export — GeoJSON (visible layers) + PNG screenshot
- Earthquake timeline — 24H/7D/30D feed with play/pause animation
- Hurricane animation — track position playback
- Keyboard shortcuts —
Llayers,?help,Escclose - Offline support — service worker with cache-first navigation
- Accessibility — ARIA roles, keyboard navigation, live regions
GPL-3.0