Question 1

What does ThermoCast predict?

Accepted Answer

Neutral thermospheric mass density in kg/m^3 at altitudes between 250 km and 600 km. Outputs are point predictions; the current model does not emit per-prediction uncertainty.

Question 2

What altitude range is supported?

Accepted Answer

250 km to 600 km in 50 km increments (250, 300, 350, 400, 450, 500, 550, 600). Queries outside this range are rejected.

Question 3

What datetime range is valid?

Accepted Answer

The valid datetime range is bounded by the model's space-weather driver dataset. Call GET /api/about to read the current first/last supported timestamps before querying a forecast endpoint.

Question 4

Which endpoint should I call?

Accepted Answer

Use /api/surface for a density map at one altitude, /api/surface/all for every altitude in one call, /api/altitude for a vertical profile at one location, and /api/density only for one-off single-point queries. Batch endpoints are more efficient than repeated single-point calls.

Question 5

How do I pin the model that produced a response?

Accepted Answer

Every API response carries X-Model-Version, X-Model-Train-Date, X-Model-File, and X-Thermocast-Version response headers. CORS exposes them, so browser clients can read them too.

Question 6

Is there a machine-readable summary for AI agents?

Accepted Answer

Yes. /llms.txt is a concise index intended for LLM agents and /llms-full.txt contains the long-form conceptual documentation. The OpenAPI specification is at /api/openapi.json.