Classification: correct Confidence: Model confidence in Bands 4 and 5 remains high. The successful prediction after recent downward adjustments to Bands 2-4 confirms the calibration is stabilizing. No further adjustments recommended.
The model correctly predicted a significant rise to 5.76 ft, which was within 8% of the actual 6.24 ft peak, despite a 4.8-hour timing error attributed to broad hydrograph shape in saturated conditions.
| Metric | Predicted | Actual | Error |
|---|---|---|---|
| Peak height | 5.76 ft | 6.24 ft | -0.48 ft |
| Total rise | — | 3.42 ft | — |
| Band | Precip | Predicted Rise | Intensity | Moisture |
|---|---|---|---|---|
| 4 | 0.96" | 0.75 ft | HEAVY | SATURATED |
| 5 | 1.27" | 0.86 ft | HEAVY | SATURATED |
The prediction magnitude was highly accurate, with only a -7.7% error (5.76 ft predicted vs 6.24 ft actual). This validates the recent calibration adjustments made on June 6th, particularly the reductions to Bands 2, 3, and 4 which previously caused false positives. The predicted contributions came primarily from Bands 4 (0.96") and 5 (1.27"), aligning with the observed QPE data where these bands had the highest accumulation during the rainfall window (13:12-15:12 UTC).
The 4.8-hour timing error (predicted 04:50 next day, actual 19:00 same day) is consistent with the 'very_broad' hydrograph shape noted in the gauge data. In saturated conditions with widespread but moderate-intensity rainfall across headwaters (Bands 4-5), the response is often slower and more dispersed than the base lag hours suggest. However, since the magnitude was correct and the event was successfully captured as a significant rise, no coefficient adjustments are warranted. The model correctly identified that this was a headwaters-driven event rather than a near-gauge burst.
No changes made.
Model confidence in Bands 4 and 5 remains high. The successful prediction after recent downward adjustments to Bands 2-4 confirms the calibration is stabilizing. No further adjustments recommended.