← 2026-06-22

Richland Creek — Daily Analysis: 2026-06-24

Classification: correct Confidence: Confidence remains 'medium'. The model is handling baseline conditions well after recent corrections, but lacks data from a significant, correctly predicted rising limb event to upgrade confidence.

Event Summary

The model correctly predicted a near-baseline peak of 1.49 ft on a day with negligible effective rainfall and no significant gauge rise, despite the timing discrepancy inherent to daily aggregation.

Prediction vs Actual

Metric Predicted Actual Error
Peak height 1.49 ft 1.55 ft -0.06 ft
Total rise 0.07 ft

Band Contributions

Band Zone Precip Predicted Rise Intensity Moisture

Analysis

The prediction generated a peak of 1.49 ft, which is within 3.9% of the observed daily maximum of 1.55 ft. This small error falls well within the acceptable +/-15-20% margin for a 'correct' classification. The watershed was in a WET state (2.124" 7-day rainfall), but the gauge precipitation was 0.0" and QPE inputs were minimal and fragmented, occurring primarily late in the day (15:13 to 23:13). Consequently, no significant hydrologic response was expected or observed beyond the baseline variability.

The large timing error (23.8 hours) is an artifact of comparing a predicted future peak (June 25) against an observed peak that occurred at the start of the reporting window (June 24 00:00). Since the observed rise was negligible (0.07 ft) and flat, the 'actual peak' likely represents the baseline starting condition rather than a response to today's inputs. The model correctly identified that today's rainfall inputs were insufficient to drive a meaningful rise above baseline levels.

Given the accuracy of the magnitude prediction for a quiet day, no calibration adjustments are recommended. The recent reductions in response coefficients (calibrated_10) appear to have successfully corrected the previous massive overpredictions. The system is currently stable for low-input events.

Coefficient Adjustments

No changes made.

Notes

Confidence remains 'medium'. The model is handling baseline conditions well after recent corrections, but lacks data from a significant, correctly predicted rising limb event to upgrade confidence.

← 2026-06-22