Forecasters

Forecasters contribute short-term weather predictions linked to specific locations and time windows, typically between 6 and 48 hours. Their submissions are timestamped, geotagged, and stored until the relevant time window has passed, at which point they are evaluated against observed outcomes.

Forecasts are evaluated on:

• Target precision: Did the predicted condition (e.g. rain, fog, wind) occur?

• Time relevance: Was the condition observed within the forecast window?

• Pattern consistency: Has the participant shown repeated accuracy over time?

Forecast accuracy influences a participant’s credibility weight within the system. Reliable forecasters may gain expanded roles or higher influence in predictive modeling.

Sometimes you're not describing the present. You're sensing what's about to unfold.

You leave a trace, and the system remembers—not just what happened, but who saw it coming.