Classification: correct Confidence: Model confidence remains stable. The ability to correctly predict 'no event' on days with trace/negligible rainfall is as important as predicting peaks. This event validates the current sensitivity settings for dry/normal conditions.
The model correctly predicted a near-baseline peak stage of 1.44 ft on a day with negligible rainfall and a flat hydrograph, closely matching the observed maximum of 1.53 ft.
| Metric | Predicted | Actual | Error |
|---|---|---|---|
| Peak height | 1.44 ft | 1.53 ft | -0.09 ft |
| Total rise | — | 0.09 ft | — |
| Band | Zone | Precip | Predicted Rise | Intensity | Moisture |
|---|---|---|---|---|---|
The prediction was numerically accurate, with a peak error of only -5.9% (predicted 1.44 ft vs. actual 1.53 ft). This falls well within the acceptable +/-15-20% margin for a 'correct' classification, particularly given that the event was essentially non-existent in terms of new runoff generation. The observed 'rise' of 0.09 ft is likely instrumental noise or minor baseflow fluctuation rather than a hydraulic event driven by the day's rainfall.
The QPE data shows minimal accumulation across both bands (max hourly rates well below intensity thresholds for meaningful runoff), and the gauge precipitation sensor recorded 0.0" for the 24-hour period. Under NORMAL antecedent moisture conditions, this lack of precipitation input correctly resulted in no significant predicted rise. The large timing error (23.8 hours) is an artifact of comparing a steady-state prediction to a daily max that occurred at the start of the window; it does not indicate a failure in hydrological routing since no distinct hydrograph peak formed to be timed.
Given the conservative nature of calibration adjustments and the fact that the model successfully identified a quiet day without false alarms, no changes to coefficients, lags, or tiers are warranted. The current response coefficients (1.161 for Band 1 and 0.855 for Band 2) are appropriate for this intensity of rainfall.
No changes made.
Model confidence remains stable. The ability to correctly predict 'no event' on days with trace/negligible rainfall is as important as predicting peaks. This event validates the current sensitivity settings for dry/normal conditions.