Classification: no_prediction Confidence: Model confidence remains unchanged. The failure to predict during saturated, zero-QPE conditions is a known limitation of the current trigger-based prediction logic rather than a coefficient calibration issue.
No prediction was generated today, yet a sharp 0.62 ft rise occurred in saturated conditions with zero QPE inputs and zero gauge precipitation.
| Metric | Predicted | Actual | Error |
|---|---|---|---|
| Peak height | N/A | 3.96 ft | N/A |
| Total rise | — | 0.62 ft | — |
| Band | Precip | Predicted Rise | Intensity | Moisture |
|---|---|---|---|---|
The physics-based predictor failed to generate any output for this event, likely due to the absence of QPE rainfall inputs across all bands. Despite this, the gauge recorded a significant sharp rise of 0.62 ft to a peak of 3.96 ft at the start of the day. This pattern mirrors the previous day's event, where residual runoff from saturated antecedent conditions caused a rise without new rainfall inputs.
Since no prediction was made, no coefficient adjustments can be calculated for the bands. However, this event reinforces the challenge of modeling recession limbs or residual runoff in saturated watersheds. The current system treats days with zero QPE as non-events, but the gauge data shows that saturated soils can still produce meaningful peaks from residual flow or unmeasured micro-catchments. The empirical model also issued no headline, missing this anomaly.
Given the 'no_prediction' classification, the calibration coefficients remain unchanged. This event highlights a gap in the prediction logic regarding saturated, zero-rainfall days, but does not provide data to tune the rainfall-response coefficients themselves.
No changes made.
Model confidence remains unchanged. The failure to predict during saturated, zero-QPE conditions is a known limitation of the current trigger-based prediction logic rather than a coefficient calibration issue.