Groundwater + surface-water atlas
Compare aquifer-subsidence risk, major rivers, reservoirs and metropolitan areas at the same regional scale.
Operational in this releaseWater Security & Resilience · Texas
See Texas water as one connected system.
Uplift Water Intelligence brings groundwater, rivers, reservoirs, water quality, drought, flooding, land subsidence, infrastructure and communities into one human-reviewed decision framework.
Responsible agentic AIAI coordinates. Experts decide.
The agent organizes sources, spatial analysis and uncertainty. Qualified professionals remain responsible for field, engineering, governance, permitting and investment decisions.
1unified system7evidence domains100%reviewed
Official live feed · USGS
Latest available Texas groundwater field measurements
Parameter 72019 · groundwater-level reading in feet. Values are shown as published and may be provisional. Field measurements are not continuous real-time sensors and can be delayed by agency processing.
Live integrations pending:
TWDB Groundwater Database ingestion is scheduled for Phase 01 through a validated download and normalization pipeline. The TWDB reports/download interface is not represented as a live API connection.
Integrated Water Intelligence
Below ground and above it—in one decision picture.
The original Texas Water Atlas is now the foundation of a larger operating model. Each layer answers a different question; together they show where water availability, hazards, infrastructure and community needs intersect.
Drought & scarcity
Connect groundwater decline with surface supply, vegetation stress, demand and monitoring coverage.
Data integration nextFlood & subsidence
Identify where sinking land may compound drainage and flood exposure for people and infrastructure.
InSAR + FEMA integration nextMonitoring & investment
Explain which evidence gaps matter, what should be validated next and why a location was prioritized.
Human review requiredWater-security priorityGroundwater condition+Surface-water availability+Drought & flood exposure+Infrastructure & community need−Uncertainty & constraints
Mission & objectives
Actionable water intelligence—with scientific accountability.
Texas Pilot · Scientific and product leadDr. Gael Ndi Nkwain
Uplift Water Intelligence combines hydrogeology, surface-water context, hazards, geospatial data and constrained AI to support evidence-based water-security decisions.
Water availability & scarcity
Map persistent aquifer decline, surface-water context and regions where demand, drought and limited supply converge.
Water quality & contamination
Connect available quality observations with wells, potential pollution pathways and public-health context while distinguishing screening from confirmed contamination.
Monitoring-network strategy
Identify places where observations are sparse, uncertainty is high and additional monitoring could add the greatest decision value.
Drought, flood & subsidence resilience
Evaluate where water stress, land movement, flood exposure, communities and critical infrastructure may intersect.
Water governance & investment
Translate evidence into auditable priorities for public agencies, utilities, groundwater districts, development partners and infrastructure planning.
Responsible agentic AI
Label values as measured, reported, model-derived or predicted. The agent coordinates approved tools; qualified experts retain authority.
Decision gateAgent synthesis→Evidence & uncertainty check→Hydrogeologist review→Field investigation
Interactive evidence mapGulf Coast Aquifer
ObservedWatchPriority
PANHANDLECENTRAL TEXASGULF COAST
N↑
0100 km
Selected evidenceGW-1042Brazoria County
- Annual trend
- -2.8 ft/yr
- Uncertainty
- 78%
- Monitoring coverage
- 22%
- Priority score
- 74/100
Built around buyer decisions
One evidence platform. Four urgent operating questions.
Each workflow is framed around a real institutional decision, with traceable sources, uncertainty, limitations and required professional review.
Where are compliance and monitoring gaps growing?
Track aquifer change, flag high-uncertainty areas and prioritize candidate monitoring zones.
Output · Monitoring evidence packageWhere could subsidence affect critical assets?
Screen roads, pipelines, drainage systems and flood barriers against reviewed groundwater–InSAR evidence.
Output · Asset risk contextWhere is scarcity pressure likely to intensify?
Compare decline patterns, monitoring coverage and reviewed trend scenarios for operational planning.
Output · Scarcity planning briefWhere do water stress and exposed communities converge?
Connect environmental evidence with infrastructure and community context without automating final decisions.
Output · Auditable decision reportData & provenance
Every future result will trace back to an authoritative record.
The interactive wells and scores above are clearly labeled demonstration data. The production pilot will ingest measured observations and published model inputs, preserve source dates and identifiers, and distinguish observations from predictions.
Current siteFunctional interface · Demonstration values→Next releaseVerified Texas observations · Traceable outputs
Primary observationsTWDB Groundwater Database
Well records, measured water levels, water quality, springs, and supporting site information.
TWDB ↗Spatial frameworkTWDB GIS DataOfficial major and minor aquifer boundaries and other downloadable Texas geospatial layers.
GIS data ↗Federal observationsUSGS Water Data APIsMachine-readable groundwater sites, field measurements, daily values, and monitoring metadata.
USGS ↗Model contextTWDB Groundwater ModelsPublished Groundwater Availability Model grids, properties, reports, and supporting datasets.
GAM downloads ↗Subsurface evidenceTWDB BRACS & Well LogsDigital geophysical well logs, lithologic and stratigraphic information, well construction, and brackish-aquifer characterization data.
BRACS data ↗Borehole intelligence
A borehole number is the key—not the answer.
Each identifier must connect to location, construction, geology, measurements, aquifer tests, and source documents. Porosity and permeability cannot be invented from a well number; they must be measured, test-derived, model-assigned, or explicitly predicted.
IdentityState well number · source agency · coordinates
ConstructionTotal depth · casing · screen top and bottom
GeologyLithology intervals · stratigraphic picks · faults
LogsGamma · resistivity · spontaneous potential · caliper
WaterStatic level · measurement date · water strike depth
PerformanceYield · drawdown · specific capacity · pump test
HydraulicsHydraulic conductivity · transmissivity · storage
QualitySalinity · TDS · major ions · sampling date
Porosity versus permeability
Porosity describes available void space. Permeability describes how easily connected pores or fractures transmit fluid. For groundwater decisions, hydraulic conductivity, transmissivity, storativity, well yield, and uncertainty are usually more actionable. Every value will carry a provenance label: measured, reported, model-derived, or predicted.
Fractured-bedrock potential
Where could bedrock hold and transmit groundwater?
The system will rank investigation zones—not promise water. High-potential areas are inferred where independent evidence converges and are then checked with field geophysics, local hydrogeology, and drilling.
01
Fractures & faultsMapped structures, lineaments, intersections and geologic contacts
02
Terrain positionValleys, slope breaks, depressions and topographic setting
03
Recharge pathwaysDrainage, soils, land cover, rainfall and surface infiltration
04
Weathered bedrockRegolith thickness and depth to competent rock
05
Nearby boreholesWater strikes, tested yield, drawdown and dry-hole evidence
06
Water suitabilitySalinity and chemistry constraints before a site is considered useful
Geology · soils · surface water
The map needs the material beneath the land—and the water moving across it.
Geologic formations define aquifers, confining units, fractures and soluble rocks. Soils influence infiltration and runoff. Rivers can recharge or drain an aquifer, but they can also transport contaminants and create flood and construction risks.
Formation name, geologic age, lithology, thickness, dip, contacts, faults, karst potential, weathering and aquifer/confining-unit role.
Texture, hydrologic soil group, saturated conductivity, drainage, depth, shrink–swell, available water capacity and flooding frequency.
Streams, lakes, watershed boundaries, flow direction, gaining and losing reaches, drainage density and proximity to possible pollution sources.
Elevation, slope, valley position, depressions, flood zones, erosion, access and protection of wellheads from inundation.
Is a river nearby good or bad?
It is evidence—not a decision.
The answer depends on whether the reach gains groundwater or loses water to the aquifer, whether surface water is contaminated, the depth and confinement of the target aquifer, flood risk and required sanitary setbacks.
Possible opportunity
- Recharge along a losing reach
- Alluvial sand and gravel aquifers
- Fractured or karst connections
- Shallow groundwater indicators
Possible concern
- Flooding and wellhead inundation
- Microbial or chemical contamination
- Bank erosion and unstable access
- Surface-water permitting constraints
Rocks & formationsTexas BEG geologic maps ↗SoilsUSDA SSURGO ↗Rivers & watershedsUSGS 3DHP ↗Flood hazardFEMA NFHL ↗
A river buffer alone must never be used to select or reject a well. The production workflow will evaluate hydrogeologic connection, contamination pathways, floodplain status, setbacks and field evidence together.
People-centered water decisions
Find the need, understand the risk, then investigate responsibly.
The system will combine hydrogeology with water-supply need, existing-well performance, pollution pathways, and community conditions. It will prioritize evidence collection—not automatically select a drilling location.
Combine projected water shortage, aquifer thickness, recharge, hydraulic properties, nearby well performance, access, setbacks, protected areas, and contamination exclusions.
- Output: candidate investigation zones
- Required next: geophysics, permits, test drilling and pump test
Track static water-level decline, increasing drawdown, decreasing yield or specific capacity, aging casing and screens, inactivity, subsidence context, and changing water quality.
- Output: monitor, inspect, rehabilitate or replace
- Never infer failure from one measurement
Screen TCEQ remediation and leaking-tank sites, Superfund records, regulated facilities, spills, agricultural and septic context, salinity, groundwater flow direction, and well construction.
- Output: source–pathway–receptor evidence
- Contamination screening can exclude a candidate zone
Connect TWDB water needs with Census population and housing characteristics, small and rural systems, distance to reliable supply, service gaps, and documented local priorities.
- Output: transparent community-priority indicators
- Community engagement remains required
Priority is not “where water might exist” aloneDocumented need+Hydrogeologic evidence+Existing-well condition+Community benefit−Pollution & feasibility constraints
A candidate advances only when evidence quality is sufficient and legal, environmental, engineering, public-health, and community reviews are documented.
Product strategy · no hype
Connect aquifer decline to the ground—and infrastructure—moving above it.
The strongest differentiator is groundwater–subsidence fusion. It expands the platform from a water-monitoring tool into defensible risk intelligence for groundwater districts, infrastructure owners, insurers, and municipal planners.
01 · Standout differentiator
Land-subsidence fusion
Groundwater withdrawal can compact aquifer-system sediments and lower the land surface. In the Houston–Gulf Coast region, that can compound risk to roads, pipelines, foundations, drainage, and flood defenses.
The proposed layer will combine verified groundwater-level change with public Sentinel-1 InSAR displacement evidence. It will show where declining groundwater and measured surface movement converge—without treating correlation alone as causation.
Groundwater districtsInfrastructureInsuranceMunicipal planning
Groundwater evidenceWater-level declineVerified well time series
+Satellite evidenceInSAR displacementSentinel-1 surface motion
=Reviewed outputSubsidence risk contextTrend · uncertainty · exposed assets
Predictive trend modeling
Combine historical decline with drought indicators and pumping trends to develop reviewed 1–3 year scenarios. Forecasts will carry assumptions, uncertainty bands, and validation status—not false certainty.
Automated continuous ingestion
Re-pull verified TWDB and USGS records on a controlled schedule, validate changes, and rerun priority scoring. This turns the platform from a one-time study into an operational service.
Plain-language evidence reports
Generate a concise, auditable explanation of why a well or zone was flagged, including sources, dates, methods, uncertainty, limitations, and hydrogeologist sign-off.
Honest caveatThe differentiators become credible only after the demonstration wells are replaced with validated observations.
- NowWire real TWDB data
- NextAdd InSAR subsidence fusion
- ThenAutomate ingestion and scoring
- ScaleAdd forecasts and auditable reports
Implementation decision · start here
Real TWDB records unlock every next layer.
TWDB publishes a public ArcGIS Feature Service containing selected wells, springs, water levels, water quality, and related groundwater records. The service architecture and query capability are verified; production ingestion, field normalization, and quality-control testing are the next engineering milestone.
Verified nowPublic service endpoint
ArcGIS REST · JSON/GeoJSON · WGS 84 point layer
Build nextValidated ingestion pipeline
Paging · county/aquifer filters · identifiers · units · dates
Then proveReal-data pilot
Gulf Coast wells · trend QA · hydrogeologist review
Not claimed yetProduction-ready automation
No live TWDB scoring or InSAR attribution is presented as complete.
Production transition plan
Four phases from research MVP to a defensible decision platform.
The sequence starts with verified data and one Gulf Coast Aquifer pilot. Statewide expansion follows only after technical validation, professional review, report automation and enforceable safety gates.
Phase 01Next
Data ingestion & live API integration
- Connect real observations: build pipelines to the TWDB Groundwater Database and USGS Water Data APIs for water levels, water quality and available daily or field measurements.
- Integrate subsurface records: ingest well-log metadata, lithology and geophysical profiles from TWDB BRACS.
- Standardize and normalize: validate State Well Numbers, dates, depth units, coordinates, aquifer names, source identifiers and measurement flags in Python.
Phase 02Build
Spatial infrastructure & Python GeoAI
- Publish ArcGIS feature layers: serve TWDB aquifer boundaries, wells, density, trends, monitoring gaps and provenance through queryable services.
- Build trend and uncertainty modules: implement reproducible time-series analysis—including Mann–Kendall where assumptions are appropriate—and spatial uncertainty mapping.
- Screen context together: integrate SSURGO soils, 3DHP rivers, FEMA flood hazard, TCEQ remediation sites and Census ACS.
Phase 03Validate
Model validation & report automation
- Launch one pilot: begin with the Gulf Coast Aquifer and a single, checkable decision question before statewide expansion.
- Ground-truth outputs: compare results with TWDB Groundwater Availability Models, measured wells, published studies and professional hydrogeologist review.
- Automate evidence reports: generate structured PDFs containing methods, inputs, data dates, provenance, uncertainty, limitations and reviewer sign-off.
Phase 04Govern
Safety guardrails & access controls
- Constrain agent tools: allow only approved analytical functions and label every value as measured, reported, model-derived or predicted.
- Require human review: a hydrogeologist must verify evidence before a candidate advances to geophysics, test drilling or permitting.
- Preserve accountability: record data versions, tool calls, assumptions, reviewer actions and access permissions for every decision package.
Recommended geospatial backboneArcGIS + Python GeoAI + constrained agent
ArcGIS handles authoritative web maps and feature layers. Python performs reproducible data preparation, trend analysis, and modeling. The agent selects approved tools, explains evidence, and records provenance—it does not invent measurements.
01TWDB + USGS→
02Python validation→
03ArcGIS layers→
04Agent analysis→
05Reviewed report
Constrained by design
The agent coordinates science. It does not replace it.
Every result follows a controlled chain of approved tools, uncertainty checks, provenance recording, and accountable human review.
01Question→
02Data→
03Spatial analysis→
04Uncertainty→
05Verification→
06Human review