← 2026-05-28  |  2026-05-30 →

Hailstone (Upper Buffalo) — Daily Analysis: 2026-05-29

Classification: correct Confidence: The model has now correctly classified three consecutive events, including two 'quiet' days and one significant rise. The magnitude prediction is robust across varying moisture conditions. Confidence remains 'low' until 3 medium/high confidence events are accumulated, but performance suggests an upgrade is imminent.

Event Summary

The physics model accurately predicted a 281 CFS peak following 0.65 inches of rain in wet conditions, with a magnitude error of only -6.0%.

Prediction vs Actual

Metric Predicted Actual Error
Peak CFS 264 CFS 281 CFS -17 CFS
Total rise 204.1 CFS

Band Contributions

Band Zone Precip Predicted Rise Intensity

Analysis

Today's event represents a significant validation of the physics model's magnitude prediction capability. With 0.65 inches of rainfall occurring in a WET moisture tier (multiplier 1.5), the basin responded with a 204.1 CFS rise to a peak of 281.0 CFS. The model predicted a peak of 264.0 CFS, resulting in a -6.0% error, which is well within the ±15-20% accuracy threshold for a 'correct' classification. The model correctly identified that wet antecedent conditions would amplify the response compared to previous dry/normal events.

The primary discrepancy was timing, with the actual peak occurring at 18:30 CDT while the model predicted a peak approximately 5.4 hours later (around 23:50 CDT / 04:51 UTC next day). Given the steep nature of the Boxley basin and the late-afternoon timing of the heaviest rain (0.162" at 18:05), the actual hydrograph rose faster than the model's dispersion parameter allowed. However, since magnitude is the primary safety metric for whitewater runs and the error is small, no coefficient adjustment is warranted.

No empirical forecast headline was issued today. This is consistent with the lack of a severe 'high' tier event (threshold 2000 CFS), though the rise to 281 CFS is notable for lower-tier activities. The absence of a headline is not penalized as the empirical model appears reserved for significant flood warnings.

Coefficient Adjustments

No changes made.

← 2026-05-28  |  2026-05-30 →