Data Sources
Hi3 Water aggregates water asset data from federal agencies, state databases, and community directories. Here's exactly where our data comes from and how we process it.
USGS National Water Information System (NWIS)
Federal AgencyThe USGS NWIS is the nation's principal repository of water resources data. We ingest site metadata for springs (site type SP) and wells (site type GW) including location, elevation, aquifer codes, and site status. Over 24,000 sites currently indexed.
- Springs
- Wells
- Streamflow gauges
- Groundwater monitoring sites
All 50 US states
Update FrequencyReal-time monitoring, site data updated quarterly
- Site name & ID
- Latitude / Longitude
- Elevation
- Aquifer code
- Site type
- State & County
USGS Principal Aquifers Map
Federal DatasetThe Principal Aquifers dataset provides polygon boundaries for all major aquifer systems in the United States. We store over 4,600 boundary polygons with rock type classification (sandstone, carbonate, igneous/metamorphic, sand & gravel), enabling the Aquifer Boundaries map overlay.
- Aquifer boundary polygons
- Rock type classification
- Geological formation data
Continental US, Hawaii, Puerto Rico, US Virgin Islands
Update FrequencyStatic dataset (updated 2023)
- Aquifer name & code
- Rock type
- Boundary polygon (MultiPolygon)
- Centroid coordinates
EPA Safe Drinking Water Information System (SDWIS)
Federal AgencyThe EPA SDWIS database tracks over 160,000 public water systems nationwide. We ingest system locations, water source classifications (groundwater vs surface water), population served counts, and link to violation records. Coordinates are geocoded from city/state data using the OpenStreetMap Nominatim service.
- Public water systems
- Water source type
- Population served
- Violations
All US states and territories
Update FrequencyQuarterly updates
- System name & PWSID
- Water source type
- Population served
- City / State / County
- Geocoded coordinates
FindASpring.org
Community DirectoryFindASpring is a community-maintained directory of natural springs. We scrape listings with full robots.txt compliance and a 2-second rate limit between requests. Coordinates are extracted from listing pages using multiple strategies (embedded maps, text parsing, geocoding). All scraped data is validated and deduplicated against existing records.
- Natural springs
- Community water sources
- Spring descriptions
North America (community-contributed)
Update FrequencyOngoing community submissions
- Spring name
- Description
- Coordinates
- State / Region
- Community notes
Ingestion Pipeline
Every data source goes through our 5-step pipeline before appearing on the platform.
Raw data retrieved from source API or scraped from directory
Coordinates, names, and data fields normalized and checked
Haversine distance + fuzzy name matching to prevent duplicates
Water Intelligence Score (WIS) computed across 5 dimensions
Asset saved to PostGIS database with full metadata
Data Quality Tiers
Every asset is assigned a data quality tier based on the depth and verification of available data.
Field verified + lab tested water quality
Reported data with partial verification
User-reported or community-sourced
Pending review and verification
Know a data source we should integrate? Have corrections for existing data?