Classification: correct Confidence: Model correctly identified a non-event. Confidence remains 'low' as no significant hydrological response was captured to validate the physics of rising stages.
A very quiet hydrological day with a 0.28" late-afternoon rain event resulted in negligible stream response; the physics model correctly predicted no meaningful rise, maintaining accuracy within <10% despite a timing artifact caused by the lack of signal.
| Metric | Predicted | Actual | Error |
|---|---|---|---|
| Peak CFS | 30 CFS | 33 CFS | -3 CFS |
| Total rise | — | 2.8 CFS | — |
| Band | Zone | Precip | Predicted Rise | Intensity |
|---|---|---|---|---|
The physics model predicted a peak of 30.1 CFS, while the observed gauge data showed a maximum of 32.9 CFS and a minimum of 30.1 CFS. The actual rise was only 2.8 CFS, which is well within the 'quiet day' variance expected for such a small rainfall total (0.279"). The predicted peak value (30.1 CFS) matches the daily minimum observed, indicating the model effectively predicted no hydrological response to the trace/low rain. The 8.5% error is well within the ±15-20% tolerance for a 'correct' classification, particularly because the magnitude of the difference (2.8 CFS) is trivial relative to the gauge's thresholds (low 500 CFS). The large timing error (22.9 hours) is an artifact of the model predicting a flat line or minimal rise that coincidentally aligns with the baseline flow at a different time of day than the actual noise peak. Given the small watershed and steep slopes, one might expect a faster response, but the rainfall intensity was low (mostly <0.1 in/hr except for brief spikes), and the total volume was insufficient to overcome infiltration losses in NORMAL moisture conditions.
No empirical forecast headline was issued, so that component is not assessed. The model's behavior confirms that the current response coefficient of 200 CFS/inch (scaled by moisture/intensity) is appropriate for suppressing predictions on sub-threshold rainfall events. Adjusting the coefficient downwards might risk under-predicting slightly larger events, while increasing it would create false positives for days like this. Therefore, the calibration remains stable.
No changes made.