01
Units: per-station bias models
The smallest reasoning components: a calibrator at each station that measures how the upstream forecast has been performing against observations at that location.
Overview
Weather forecasting calibrated to operational decisions.
P1-Tempo is the environmental-intelligence system on the Intelligence OS stack. Per-station bias models correct the upstream forecast; the calibration pipeline produces a 30-day probabilistic field. Chaos is a Function within Tempo. Tempo is a governed System in its own right, and a governed input to Maritimo and Terra. The result is a forecast that is always the current one, corrected against the observations present at the moment it is issued.
CURRENT RESULTS ARE A BASELINE, NOT AN UPPER BOUND.
The field, not the issuance.
A weather forecast is traditionally issued at a fixed cadence and held as equally good at every station until the next run. Real error is local. A model that is reliable in aggregate can be systematically off at the point a ship, a farm, or a corridor actually decides.
Tempo treats each station as its own forecasting problem. Bias models run independently, refreshed against observations, and the field they produce is what Maritimo and Terra optimise over. Where a regional forecast ends, Tempo begins.
In practice
In practice
02
Functions: calibration pipeline
Declared relations between models. Chaos supplies the ensemble; Tempo corrects it. Signal is extracted, not averaged away.
03
System: 30-day probabilistic forecast
A governed System in its own right, and a governed input to Maritimo and Terra. Operable at either level of granularity.
Position in the Principia stack
Tempo as an Intelligence OS application
Tempo is a domain-specific instantiation of the Intelligence OS architecture applied to environmental intelligence. Calibration precision at each station determines the quality of the field Maritimo and Terra reason over, which determines the route, the season, and the realised outcome. Forecasts are traceable to the station, the bias model, and the cycle that produced them.
Next
Talk to P1 Tempo
Tell us which stations and corridors you already run. We will come back on the calibrated field they produce.