USGS
Hydrology and groundwater observations (station/site level).
Section 01
DATA / SEE THE EVIDENCE, NOT JUST THE STORY
NOAA
Drought and climate history for trend context.
USDA + CENSUS
Irrigation demand and demographic pressure overlays.
Section 02
FLAGSHIP METRICS (WITH SOURCE PATHS)
AQUIFER + SUPPLY PRESSURE
- Ogallala: 332 MAF depleted since 1950 (20% of 2,925 MAF total). GRACE satellite: -294 km³ (2nd globally). National depletion: 1,000 km³ since 1900.
- Crisis probabilities by 2100: Edwards 53.2% [52.7%, 53.6%], Floridan 36.3% [35.9%, 36.7%], Central Valley 17.3% [16.9%, 17.6%].
- Combined Dust Bowl probability: 29.9% [29.5%, 30.3%]. Combined megadrought: 88.4%. 24.5M Americans at risk.
Sources: FINDINGS.md, 06-data-provenance-and-methods.md.
STORMWATER OPPORTUNITY
- Total stormwater available: ~29.8 MAF/yr across 8 major aquifer regions (USGS streamflow + NOAA precip normals).
- ASR capture potential: ~20.8 MAF/yr with 70% capture efficiency (from water budget analysis).
- 6 of 8 regions show EXCELLENT ASR viability. Central Valley (MODERATE) most constrained; Basin & Range already in surplus.
Sources: FINDINGS.md, 05-engineering-masterplan.md.
Section 03
CHART/INFOGRAPHIC BLUEPRINTS
CHART 1
Aquifer depletion by basin with current estimate + critical-year range.
CHART 2
Stormwater availability vs current capture vs modeled capture capacity.
CHART 3
Megadrought probability timeline with P10/P50/P90 uncertainty bands.
CHART 4
Do-nothing losses vs phased ASR build costs over 25 years.
Section 04
INTERACTIVE PROVENANCE EXPLORER
Data Provenance Explorer
Trace Every Key Number to Source + Method
Primary source
USGS NWIS groundwater level records
Computation method
Filtered by aquifer + time-series completeness, 1980-2025. Theil-Sen regression with bootstrap 95% CI on all slopes.
Key result
Top urgency: Central Valley, CA (score 85.2, 81% wells declining)
Section 05
WATER BUDGET BALANCE SHEET
Section 06
PROVENANCE RAIL
Claims are labeled as measured, derived, or modeled. This is deliberate: investors and policy teams need to see where confidence is high, where assumptions dominate, and where additional field verification is required.