Day 21 (2026-03-21) — Continued Extended Dry Period
Zero precipitation across all 37 HUC12 sub-watersheds for the third consecutive day. This extends the current dry spell to 8 consecutive zero-precipitation days (since 3/14). The 7-day antecedent precipitation remains unchanged from Day 20 at 0.14-0.56" across the watershed — the driest sustained period of the study.
All gauges continue recession. No significant rises detected at any gauge. End-of-day values and daily changes:
| Gauge | Day 20 EOD | Day 21 EOD | Change | Day 21 Rate (est.) |
|---|---|---|---|---|
| Boxley | 2.16 ft | 2.12 ft | -0.04 ft | Very slow decline |
| Ponca | 129 cfs | 122 cfs | -7 cfs (-5.4%) | ~0.23%/hr |
| Pruitt | 153 cfs / 3.96 ft | 140 cfs / 3.91 ft | -13 cfs (-8.5%) / -0.05 ft | ~0.38%/hr |
| St. Joe | 353 cfs / 4.02 ft | 324 cfs / 3.94 ft | -29 cfs (-8.2%) / -0.08 ft | ~0.36%/hr |
| Harriet | 425 cfs / 4.04 ft | 385 cfs / 3.97 ft | -40 cfs (-9.4%) / -0.07 ft | ~0.42%/hr |
| Richland | 1.16 ft | 1.18 ft* | +0.02 ft | See anomaly note |
| Bear Creek | 2.39 ft / 10.6 cfs | 2.37 ft / 9.3 cfs | -0.02 ft / -1.3 cfs | Approaching floor |
Recreational status: - Ponca: 122 cfs — Too Low (<150). Crossed below 125 around 09:00 CST. Deepening into Too Low range. - Pruitt: 140 cfs — Low but Floatable (100-200) per threshold, but likely reading low per local knowledge about gauge scouring. Dropped below 150 by mid-afternoon. - St. Joe: 324 cfs — Optimal (200-8000). Still well within Optimal range. - Harriet: 385 cfs — Optimal (200-9370). Still well within Optimal range.
None — no precipitation to pair.
No propagation events to track.
Continued post-Event 3 recession, now Day 10 since last significant rain (Event 3 on 3/11-12). Key recession observations:
Ponca baseflow convergence: Ponca dropped from 129→122 cfs today, a rate of ~0.23%/hr. On Day 20 the rate was ~0.21%/hr. The deceleration observed on Day 20 appears to have paused or slightly reversed — the gauge dropped 7 cfs today vs. 7 cfs yesterday (129-122 vs. 136-129). This suggests we have NOT yet reached the baseflow floor. Revising the Ponca baseflow floor estimate downward to ~115-120 cfs. The previous estimate of 120-125 cfs was premature.
Pruitt low-flow behavior: Pruitt CFS values today show extreme noise (range 133-155 cfs within the day, an 18% spread). Height is steadier (3.88-3.97, 0.09 ft range). The height data show a cleaner trend of continued slow decline. The CFS noise at these low stages further supports the local knowledge about the gauge reading unreliably at low water. The Pruitt-Ponca gain today: ~140 - 122 = 18 cfs (15%), continuing the collapse from 61→41→24→18 cfs over 4 days. This is almost certainly a gauge artifact rather than a real disappearance of the Pruitt zone's baseflow contribution.
St. Joe recession: Dropped from 353→324 cfs, rate ~0.36%/hr. This is slightly slower than yesterday's 0.43%/hr, suggesting continued deceleration. The large drainage area (1,932 km², 14 HUC12s) is sustaining flow well but still actively declining.
Harriet recession: Dropped from 425→385 cfs, rate ~0.42%/hr — essentially unchanged from Day 20. The Harriet gauge shows remarkably steady recession behavior. The St. Joe-to-Harriet gain is now 385-324 = 61 cfs, contributed by Bear Creek (9.3 cfs) plus Harriet zone (5 HUC12s, 430 km²) = ~52 cfs from the Harriet zone proper. This is consistent with a large baseflow contribution from the lower watershed.
Richland Creek mid-day spike: Richland jumped from 1.14-1.15 ft baseline to 1.20 ft around 12:45-13:15 CST, then slowly declined through the afternoon. This occurred with zero MRMS precipitation in the Richland zone. The 0.06 ft hourly rise flagged by the gauge summary is real in the data. Possible explanations: 1. Spring discharge pulse — karst springs in the Richland watershed could produce episodic flow increases unrelated to same-day precipitation. 2. Very localized rainfall missed by MRMS — unlikely given the 1-km resolution would catch most events, but possible for very small-scale convection. 3. Gauge artifact/debris — a brief obstruction could cause a step change.
The most parsimonious explanation is karst spring activity. This is the first observation of this type in the study and worth monitoring. If it recurs on subsequent dry days, it would confirm a karst discharge mechanism in the Richland basin. This is relevant to understanding the Richland Creek detection thresholds — the gauge has a noise floor that includes karst spring variability of ~0.05-0.06 ft even with no rainfall.