← All Research ResourcesmdSource: USGS National Water Information System (NWIS) daily streamflow records For each region: identify gauging stations in tributary watersheds overlying target aquifers Calculate flow duration curves (exceedance percentiles) from 30+ year records Define "surplus threshold" as flow exceeding the 25th percentile (i.e., flows exceeded 75% of the time are protected for downstream/ecological use) **Environmental flow reserve:** No diversion below 7Q10 (7-day, 10-year low flow) + 20% safety margin **Legal water rights:** Diversion only from unappropriated flows or junior-right surplus **Seasonal windows:** Capture concentrated in wet season (varies by region: Nov-Apr in West, Mar-Jun in Plains) **Fail-closed design:** If any constraint is violated, diversion stops automatically Pre-treatment screening losses: 2-5% Treatment reject/waste stream: 5-15% (varies by source water quality class) Injection well efficiency: 85-95% (accounts for clogging, backflush, and maintenance downtime) Net recovery efficiency: 70-85% of injected volume (aquifer-dependent)
Climate non-stationarity: historical streamflow records may not represent future conditions Practical capture rates depend on infrastructure buildout pace and permitting timelines Aquifer injection capacity varies significantly by geology (alluvial >> consolidated rock) The "190 trillion gallons" figure represents total stormwater runoff in aquifer regions; practical capture is 10-30 trillion gallons (5-15%) given physical and regulatory constraints USGS NWIS: https://waterdata.usgs.gov/nwis USGS Groundwater Decline and Depletion: https://www.usgs.gov/mission-areas/water-resources/science/groundwater-decline-and-depletion State water allocation databases (varies by state)
Data Provenance
Capture Model Method
Capture Model Method
Overview
The stormwater capture model estimates annual recoverable volume from flood-surplus diversions across five major aquifer regions.
Methodology
1. Streamflow Exceedance Analysis
2. Diversion Constraints
3. Treatment Loss Factors
4. Regional Capture Estimates
| Region | Annual Surplus (MAF) | Practical Capture (MAF) | Notes |
|---|---|---|---|
| Ogallala | 8-12 | 2-4 | Limited by distance from major rivers; relies on Platte, Arkansas, Canadian tributaries |
| Central Valley | 5-15 | 2-5 | Highly variable; flood years can be 5x dry years; Sacramento/San Joaquin system |
| Basin & Range | 3-8 | 1-3 | Ephemeral washes, monsoon-driven; Phoenix, Tucson, Las Vegas corridors |
| Edwards | 2-4 | 1-2 | Guadalupe, San Antonio, Nueces rivers; well-characterized karst recharge zones |
| Denver Basin | 1-3 | 0.5-1.5 | South Platte, Arkansas headwaters; limited by seasonal window |
5. Uncertainty and Caveats
Data Sources
Classification
Modeled — Based on measured streamflow data with engineering assumptions for capture efficiency.
Raw file: /research/methods/capture-model.md