FE / DATA
Public satellite oceanography.
Named sources, honest timestamps.
FishEdge is built on open data from NOAA, NASA, and Copernicus. Every layer on the map is labeled with where it came from and when it was valid. If a layer is missing, you see the gap, not a fabricated fill.
| Data product | Status | Source | Refresh |
|---|---|---|---|
| Sea-surface temperature | Live | NOAA Geo-Polar Blended SST | 6 hours |
| Chlorophyll-a concentration | Live | NASA MODIS / VIIRS / OLCI | Daily |
| Surface currents | Live | Copernicus GLOBAL_ANALYSIS_FORECAST | 12 hours |
| Sea-surface height anomaly | Beta | Copernicus satellite altimetry | Daily |
| Thermal front detection | Beta | Derived from SST gradient analysis | 6 hours |
| Confidence + evidence chain | Roadmap | Composite of the above | — |
| Route export (GPX) | Roadmap | — | — |
| Trip forecasts (multi-day) | Roadmap | Derived from NCEP / ECMWF | — |
What "live" actually means
Satellite data is delayed by orbit, processing, and ingest. FishEdge refreshes the dashboard every six hours from the latest available pass. The timestamp on every layer tells you when that pass was — not when the dashboard reloaded.
Why these three
Temperature, color, and currents are the three signals captains have historically read together from separate tools. FishEdge puts them on one map with the same coordinate system, same timestamp, and same zoom. The intent is to shorten the time between opening a chart and forming a working thesis about where to look tomorrow.
What's not in here yet
- Confidence scores — multi-signal synthesis that combines temperature, color, and current agreement into a single recommendation. In research. Not shipped.
- Forecast horizons — multi-day forward projections. Roadmap, awaiting validation.
- Auto-generated plans — anything that selects a starting position for the captain. Explicitly out of scope for the current product.
Coverage limits
The dashboard currently covers the U.S. Mid-Atlantic Bight (Delmarva, Virginia, the Carolinas). Coverage expands as the data pipeline and feedback loops support it. There is no fabricated fallback — if a region has no fresh data, the map says so.