Updated Hypothesis Document

Buffalo River Watershed — Running Hypothesis Document

Last updated: 2026-03-11 (Day 11) Events observed to date: 3 confirmed rainfall-runoff events - Event 1: March 5-6, 2026 — moderate upper watershed storm (Tier 2) - Event 2: March 7-8, 2026 — major widespread storm (Tier 3, High Water) — COMPLETE, recession interrupted by Event 3 - Event 3: March 11, 2026 — moderate upper watershed storm (Tier 2) — IN PROGRESS, wave propagating toward St. Joe

Study Context

This is the hypothesis document for a 90-day empirical study (March–June 2026) correlating MRMS radar-estimated precipitation with USGS gauge responses across the Buffalo River watershed (HUC8 11010005, 3,471 km²). Seven active gauges monitor five mainstem points and two tributaries. The watershed is divided into 37 HUC12 sub-watersheds, 29 of which drain to gauged locations (79.9% of total area).

All findings below marked "prior knowledge" come from paddler experience and existing creek monitoring — they are starting hypotheses, not empirical results from this study. Entries marked "observed" come from study data.

EVENT LOG

Event 1: March 5-6, 2026 — Upper Watershed Storm

• Classification: Tier 2 (Recreational response) — CONFIRMED
• Precipitation: Widespread 0.4-1.0" across watershed; concentrated 1.038" in Boxley headwaters with 0.751"/hr peak intensity.
• Primary gauges affected: Boxley +0.87 ft; Ponca +95 cfs (83.7→179); Pruitt +0.71 ft / +144 cfs (62→206); Richland +0.26 ft; Harriet +0.04 ft
• Gauges with no significant response: St. Joe, Bear Creek
• Propagation: Boxley → Ponca (4-5 hr) → Pruitt (20 hr from Boxley peak) at low flow velocities
• Recreational impact: Pruitt crossed into Optimal (>200 cfs) for ~2.5 hours on 3/06

Event 2: March 7-8, 2026 — Major Widespread Storm (COMPLETE — recession interrupted by Event 3 on Day 4)

• Classification: Tier 3 (High Water) — CONFIRMED
• Precipitation: Widespread 1.0-2.5" across entire watershed. Peak zones: Whiteley Creek 2.545" (0.961"/hr), Cove Creek 2.43" (0.74"/hr), Beech Creek 1.964" (0.768"/hr). Zone averages: Pruitt 2.022", Ponca 1.918", St. Joe 1.358", Richland 1.24", Bear Creek 1.199", Boxley 1.135", Harriet 1.071".
• Antecedent conditions: Moderately wet from Event 1 (0.7-1.45" 7-day). This amplified responses dramatically.
• Complete gauge response summary:

• Recession: Tracked through Day 3 (3/10) cleanly, then interrupted by Event 3 on 3/11. Pre-Event 3 levels: Ponca 197 cfs, Pruitt 315 cfs, St. Joe 891 cfs, Harriet 1,130 cfs, Boxley 2.63 ft.

Event 3: March 11, 2026 — Moderate Upper Watershed Storm (IN PROGRESS)

• Classification: Tier 2 (Recreational response) — CONFIRMED at upper gauges; wave propagating downstream
• Precipitation: Concentrated in upper watershed: Ponca zone 0.826" (peak Smith Creek 0.984", 0.620"/hr), Boxley 0.811" (0.390"/hr), Pruitt zone 0.526" (0.254"/hr), St. Joe zone 0.505" (distributed), Richland 0.565", Bear Creek 0.518", Harriet 0.333".
• Antecedent conditions: Very wet — 1.5-3.3" 7-day precip. Gauges still elevated from Event 2. This is a stacked event.
• Gauge response summary (as of EOD 3/11):

• Signal separation: Richland showed only marginal response (0.57" rain → 0.08 ft rise). Bear Creek showed no response (0.52" rain → nothing). Both confirm detection thresholds.
• Expected 3/12: St. Joe should peak ~04:00-08:00 CST 3/12. Harriet will see wave arrival later on 3/12.
• Recreational impact: Extended Optimal conditions at Ponca and Pruitt that were about to end from Event 2 recession. Ponca re-entered Optimal range decisively. Pruitt was boosted from ~315 cfs back to 857 cfs.
• Boxley recreation note: Reached 3.28 ft — briefly touched the Low-but-Floatable threshold for Boxley-to-Ponca section (3.2 ft per local knowledge) for ~2 hours. Did not reach Hailstone threshold (3.7 ft).

Propagation Timing (Observed — UPDATED with Event 3)

Event 1 (Low Flow — ~150 cfs average transit conditions):

Event 2 (Flood Flow — ~1,500-3,000 cfs transit conditions):

Event 3 (Moderate Flow — ~400-800 cfs transit conditions):

Flow-Dependent Propagation Velocity — Ponca→Pruitt Reach (KEY UPDATE — 3 events):

Velocity-discharge relationship (Ponca→Pruitt): Three data points now allow fitting. A power law V = a·Q^b with the three points suggests V ≈ 0.06 · Q^0.35 (V in m/s, Q in cfs). This fits better than √Q. At 1,000 cfs, this predicts ~0.93 m/s (~6.5 hr lag). At 3,000 cfs, it predicts ~1.3 m/s (~5.3 hr lag). Confidence: Moderate (3 events, consistent trend). Need more data points in the 500-1,500 cfs range.

Flood Wave "First Arrival" vs Peak (NEW from Event 3):

At St. Joe, the leading edge of the Pruitt wave (recession arrest/slight uptick) was detected ~5 hours after Pruitt peaked, while the crest is expected ~12-16 hours after Pruitt peak. This means the first signal arrives ~7-11 hours before the crest. For flood forecasting, this provides substantial early warning at St. Joe once a rise is confirmed at Pruitt.

Flood Wave Attenuation (from Event 2):

Key finding: At these flow levels (~3,000 cfs), channel