Classification: correct Confidence: The model has now had 6 events since the last major calibration. Magnitude accuracy is high. Confidence can be upgraded to 'medium' as we have surpassed the minimum event threshold for medium confidence and the predictions are consistently accurate in magnitude.
The physics model accurately predicted the peak flow of 325 CFS following 0.66 inches of rainfall in wet antecedent conditions, with a negligible magnitude error of -1.8% despite a significant timing lag.
| Metric | Predicted | Actual | Error |
|---|---|---|---|
| Peak CFS | 319 CFS | 325 CFS | -6 CFS |
| Total rise | — | 236.7 CFS | — |
| Band | Zone | Precip | Predicted Rise | Intensity |
|---|---|---|---|---|
The physics model performed exceptionally well in terms of magnitude, predicting a peak of 319 CFS against an actual observed peak of 325 CFS, resulting in an error of only -1.8%. This is well within the ±15-20% threshold for a 'correct' classification. The model successfully captured the high runoff efficiency of the Boxley basin under wet conditions (2.35" 7-day total), driven by the late-day burst of rainfall (0.291" at 20:05). The previous 15% upward adjustment to the response coefficient appears to have been appropriate, as the model no longer underestimates peak flows as significantly as it did in the previous event.
However, the timing prediction remains a persistent issue. The predicted peak occurred at 2026-06-08T04:51 UTC (23:51 CDT on June 7), while the actual peak was observed at 22:15 CDT on June 7. This results in a timing error of approximately 1.6 hours, with the model predicting the peak later than it actually occurred. Given the steepness of the watershed and the rapid response characteristics observed in recent history, the current lag/dispersion settings may be slightly too long for this specific intensity profile.
No empirical forecast headline was issued for the hailstone gauge today. Therefore, the empirical model's accuracy cannot be assessed for this specific event. The physics model's performance here reinforces the recent calibration trend, suggesting the current response coefficient of 230 CFS/inch (after the +15% increase) is robust for wet conditions. Further adjustments to the magnitude coefficient are not warranted given the high accuracy.
No changes made.