Classification: false_negative Confidence: Band 1 confidence remains medium due to persistent issues with unmeasured near-gauge rainfall causing false negatives. Other bands remain consistent.
The model predicted a negligible rise (3.26 ft) based on minimal QPE, but the gauge recorded a significant 0.38 ft sharp rise to 3.63 ft at the start of the day, indicating missed near-gauge rainfall or unmodeled antecedent moisture contribution.
| Metric | Predicted | Actual | Error |
|---|---|---|---|
| Peak height | 3.26 ft | 3.63 ft | -0.37 ft |
| Total rise | — | 0.38 ft | — |
| Band | Precip | Predicted Rise | Intensity | Moisture |
|---|---|---|---|---|
| 4 | 0.13" | 0.06 ft | LIGHT | WET |
| 5 | 0.09" | 0.03 ft | LIGHT | WET |
The prediction engine failed to capture the magnitude of the rise occurring at the start of the analysis period. The predicted peak of 3.26 ft underestimated the actual peak of 3.63 ft by 0.37 ft (-10.2%). More critically, the timing error of nearly 24 hours suggests the model identified the wrong hydrograph phase entirely; the actual rise occurred at 00:00, while the model predicted a response to rainfall that didn't occur until later in the afternoon (14:12-15:12) which was too small to cause such a rise. The gauge recorded 0" precipitation, yet a sharp rise of 0.38 ft occurred. This pattern mirrors the 'hidden signal' issue noted in Band 1 calibration history, where near-gauge rainfall or rapid saturation response is not captured by QPE or gauge sensors.
| Band | Change | Reason |
|---|---|---|
| 1 | +20% | The sharp rise at 00:00 with 0" gauge rain and minimal QPE in Bands 4-5 at that time strongly suggests unmeasured near-gauge rainfall (Band 1) or immediate response from wet soils not captured by current coefficients. Previous events with similar signatures required upward adjustments to Band 1. |
Band 1 confidence remains medium due to persistent issues with unmeasured near-gauge rainfall causing false negatives. Other bands remain consistent.