Classification: no_prediction Confidence: Confidence remains mixed due to recent false positives and non-generative events. The model is currently under-predicting (or not predicting) non-rainfall driven fluctuations, but this is an acceptable failure mode for a rainfall-driven model.
The prediction engine failed to generate output for a day with a sharp, unexplained 0.31 ft rise at the gauge, despite zero QPE and zero recorded precipitation.
| Metric | Predicted | Actual | Error |
|---|---|---|---|
| Peak height | N/A | 3.33 ft | N/A |
| Total rise | — | 0.31 ft | — |
| Band | Precip | Predicted Rise | Intensity | Moisture |
|---|---|---|---|---|
The prediction system returned no peak value today, classifying this as a 'no_prediction' event. This is the third consecutive day of non-generation following a period of significant activity and recent coefficient reductions. The gauge data shows a sharp 0.31 ft rise occurring immediately at the start of the day (00:00), returning from 3.02 ft to 3.33 ft. This rise shape and timing are highly atypical for rainfall-driven runoff given that all QPE bands show 0.00" and the gauge precipitation sensor recorded 0".
This mirrors the behavior seen on June 8 and June 9, where sharp rises occurred with no precipitation inputs. These events likely represent residual drainage or tailwater effects from the previous saturated event (June 6-7) rather than new rainfall response. Since the predictor correctly generated no output for a day with zero antecedent rainfall triggers, and the observed rise is not rainfall-driven, no calibration adjustments are warranted. The current coefficients appear appropriately conservative for zero-rainfall days.
No changes made.
Confidence remains mixed due to recent false positives and non-generative events. The model is currently under-predicting (or not predicting) non-rainfall driven fluctuations, but this is an acceptable failure mode for a rainfall-driven model.