See the system. Build the data model. Then simulate.
DataNET is MAGNET4WATER’s spatial intelligence layer: a live, global, multiscale environment for discovering, fusing, visualizing, transferring, and federating water and environmental data.
Before assumptions are made and before a physics model is built, users can draw a box, create a default data-fusion model, perform a virtual site visit, see the unseen, characterize the system, and decide what needs deeper modeling.
Live geospatial services from USGS, NASA, NOAA, ESA, USDA, EPA, the British Geological Survey (BGS), the International Groundwater Resources Assessment Centre (IGRAC), the Multi-Resolution Land Characteristics Consortium (MRLC), Stanford, the Cornell University Geospatial Information Repository (CUGIR), the Flanders Marine Institute (VLIZ), and many more — federated through OGC WMS / WFS / WCS and ArcGIS REST. Live, queryable, model-ready.
See DataNET turn the world's networked data into a working model of the water system.
See the data-first paradigm in action: federated services, telemetry, 2D/3D views, transfer into MAGNET platforms.
Standing on the shoulders of giants. The data is theirs.
DataNET doesn't generate data — it stands on the agencies and programs that do. National hydrological networks, geological surveys, space programs, soils science programs, and international data initiatives have spent decades building the observational foundation. HydroSimulatics' contribution is twofold: (1) we publish many agency datasets as web services on our own GeoServer (~60,000 services, converted from shapefiles, GeoTIFFs, and CSVs into WMS/WFS/WCS) so that data which would otherwise stay in static downloads becomes consumable across the world; and (2) DataNET federates HydroSimulatics' GeoServer alongside many other agency portals worldwide, making the whole ecosystem reachable through one interface. Either way, the data is theirs — and so is the science behind it.
Twelve foundational data programs.
USGS
The single largest source of US water and earth-observation data. NWIS streamflow, 3DEP terrain, NHDPlus hydrography, well records, lithologic and geological data. ScienceBase and data.gov host much of it; we've republished much of it as services.
NASA
The global Earth-observation record, often the gold standard. Landsat (50-year imagery), MODIS, SMAP soil moisture, GRACE / GRACE-FO groundwater storage, LDAS land assimilation, SEDAC socio-environmental.
NOAA
The climate and weather backbone. PRISM 4km, CFSR reanalysis, Atlas 14 precipitation frequencies, NEXRAD radar, NWS forecasts, NWM streamflow. Foundational for any continuous-time water model.
USDA / NRCS
US soils science. SSURGO 30m, STATSGO2, gNATSGO 10m where available. The parametric foundation for every US groundwater and watershed model that needs hydraulic conductivity, soil texture, runoff curve numbers.
EPA
US water quality and compliance. WATERS, STORET, ECHO. Regulatory boundaries, monitoring records, enforcement data — what makes water-quality modeling defensible against regulation.
ESA / Copernicus
The European satellite program. Sentinel-1 InSAR (subsidence and surface motion), Sentinel-2 multispectral, Sentinel-3 ocean, Sentinel-5P atmospheric. Global, free, continuously expanding.
BGS · United Kingdom
British Geological Survey. Borehole records, geological maps, aquifer characterization, groundwater monitoring across the UK and partner regions. One of the world's most comprehensive national subsurface archives.
BRGM · France
Bureau de Recherches Géologiques et Minières. ADES groundwater levels, BSS borehole database, geological maps and hydrogeology. The French national geological survey and a major contributor to European groundwater science.
BGR · Germany
Bundesanstalt für Geowissenschaften und Rohstoffe. Geology, hydrogeology, soils, mineral resources at federal scale. Substantial contributor to international hydrogeological assessments (IHME, WHYMAP).
Geoscience Australia
Australia's national geoscience agency. Continental geological mapping, water resources, terrain, satellite-derived layers. Major contributor of original data we publish on our own GeoServer.
FAO
UN Food and Agriculture Organization. Harmonized World Soil Database (HWSD), global soils, agricultural water statistics, transboundary water assessments. The standard for global agricultural and water-resource work.
IGRAC · UNESCO
International Groundwater Resources Assessment Centre. Transboundary aquifers, the Global Groundwater Monitoring Network (GGMN), country profiles. The UNESCO-hosted resource for global and cross-border groundwater work.
Two paths to the same data — convert, or federate directly.
HydroSimulatics' GeoServer (one node in DataNET's federation)
~60,000 converted services. Built from the agencies above by taking their shapefiles, GeoTIFFs, and CSVs and converting them into WMS/WFS/WCS services on HydroSimulatics' GeoServer. Some of these services back the preprocessed multi-resolution database that direct-links to the four modeling platforms (separately from DataNET); the rest are reachable through DataNET like any other federated node. The engineering work that makes formerly-static agency data consumable as live services across the world.
Direct federation via OGC standards
~14,000 layers across 145+ external provider domains. DataNET reaches into agencies' own portals (USGS ScienceBase, NASA Earthdata, BGS Maps, BRGM InfoTerre, Geoscience Australia services, plus university repositories like Stanford GeoWebServices, Cornell CUGIR, INSPIRE EU portals, VLIZ Marine, GNS Science, U. Minnesota uSpatial, and many more) via WMS, WFS, WCS, and ArcGIS REST. The federation that keeps DataNET current as agencies update their services.
The standards and software that made federation possible.
Standards bodies built the protocols: OGC (WMS, WFS, WCS, OGC API, CSW, WMTS), W3C and IETF (HTTP, REST, JSON, GeoJSON). Cloud-native formats make petabyte archives queryable by geometry: STAC, COG, Zarr, Parquet. Open software is the engineering backbone: GeoServer, MapServer, THREDDS, ERDDAP, PostGIS on the data side; Cesium, VTK, deck.gl, MapLibre, Leaflet, OpenLayers on the rendering side. Without these, federation would be impossible.
The data is already networked. Build your model on the network — and make it accessible.
Don't download. Build on the network.
The world's water data is not in silos. It is already networked — published by USGS, NOAA, NASA, and equivalents worldwide through standardized web services (WMS/WFS/WCS). The architecture is there. The data is there. What's missing is the workflow that uses it on the network. Conventional workflows download the data from the network into project-by-project silos, then process it locally, then publish results no one else can find. Every project rebuilds what already exists. The network gets broken at the moment of download. And the data is increasingly immovable — LiDAR at meter scale across continents, satellite reanalysis at sub-hourly cadence, soils at 10m globally. You cannot download what is already too large to move. DataNET inverts the workflow: the data stays on the network, the model is built on the network, and the four physics platforms (IGW-NET, SwaNET, StormNET, ConduitNET) are reached when process-based simulation is needed.
Characterization is already a model.
Every water project begins by fusing heterogeneous evidence — terrain, wells, lithology, monitoring records, land use, imagery, historical reports, local knowledge. DataNET formalizes this work as the first model. The act of selecting, fusing, rendering, and interpreting data is itself a form of modeling: a well-built spatial characterization is not preparation for the real model — it is a defensible model of the system. Before equations, evidence. Before assumptions, the data. Before simulation, characterization.
Too little direct data. Too much controlling data.
Water decisions face a paradox no amount of additional data alone can resolve. There is too little direct data — the state variables of direct interest (flow, head, water quality, concentration, flooding, system performance) are sparse at the locations, scales, and time resolutions decisions require. And there is too much controlling data — the spatial datasets that govern system behavior (terrain, soils, land use, hydrography, climate, infrastructure) exist at massive resolution, are increasingly seamless globally, and are increasingly immovable. Sparse measurements tell us what is happening. Dense controlling data tells us why. Both problems are real; neither alone solves the other.
Use what controls. Calibrate with what's measured. Refine with what models reveal.
The resolution is fusion. DataNET connects users to a federated network of authoritative data services — agency observations, monitoring networks, geospatial publications, telemetry, remote sensing, lithology and well records, water-quality samples, and HydroSimulatics-published services — and lets the user assemble them into one site characterization that brings sparse direct measurements together with dense controlling data into a defensible system understanding.
DataNET is the first step in MAGNET.
Such a federated, multiscale, observation-driven characterization environment does not exist as established practice today. The components exist — agency data, web services, cloud computing, process models — each within its own portal, format, or workflow. What's not yet established is the architectural integration. DataNET is that integrated characterization environment: the layer that turns scattered data into system perception and formalizes site characterization as a systematic, repeatable workflow.
Spatial reasoning is the working language.
Data-driven modeling fits how water-systems work actually gets done. Planners read overlays. Utility managers reason from infrastructure maps. Regulators interpret monitoring records spatially. Hydrogeologists synthesize bedrock geology, wells, water quality, and ecological indicators to see flow paths. The work has always been spatial reasoning across heterogeneous evidence — DataNET formalizes the workflow and gives every layer a sourced, auditable origin. The cognitive load of fusion is the cognitive load of professional spatial reasoning: stack the layers, see the overlays, identify the relationships, build the defensible story. The workspace is a defensible exhibit — every layer is a sourced public service, auditable by anyone who opens it.
The four physics platforms make process-based modeling dramatically more accessible to the technical community that needs it: engineers, scientists, consultants, regulators using PDE-based simulation. They take what used to require batch jobs, format wrangling, and post-processing tools and deliver working models in minute ten. DataNET reaches the larger professional community whose work runs in spatial reasoning across evidence physics cannot ingest — the same professionals named above, joined by journalists, educators, and ecologists who synthesize indirect evidence daily. Two paradigms, two professional communities, one networked data architecture — both producing rigorous water-systems understanding in their own working languages.
Each paradigm sees a different cone of evidence.
Process-based models see the variables that appear in their governing equations. Data-driven models see everything that geo-references to the same place — including the indirect evidence that often makes the difference between knowing a water system and merely simulating part of it.
Process-based equations have a fixed appetite. Data-driven fusion does not.
Process-based models can only ingest what their equations allow. An equation responds to a variable only if that variable appears in the equation. A groundwater flow simulation ingests recharge, hydraulic conductivity, boundary heads — not eco-indicator surveys or photographic surveys of stream channels. A surface-water routing model ingests precipitation, land cover, channel geometry — not lithology or contamination plume observations. This is not a software limitation; it is what physics-based modeling is. Each process-based model is constrained to the data its governing equations need.
Data-driven models have no such constraint. Fusion can ingest anything that geo-references to the same place — observations the physics-based model cannot use, soft and indirect data that still tells the analyst something, qualitative records, historical evidence. A data-driven model of a watershed can include both the hydraulic conductivity field that an IGW-NET model uses AND the historical land-management records that no physics equation accepts. The fusion sees more than any single physics-based model can.
Indirect evidence is often the most diagnostic.
Direct measurements of water-system state — heads, flows, age tracers, contaminant concentrations, pumping tests — are expensive and sparse. Indirect evidence is abundant and often free, because someone else collected it for a different purpose: satellites already image temperature, agencies already publish lithology and land use, fish surveys already record cold-water reaches, and InSAR has been measuring subsidence everywhere since Sentinel-1 launched. Much of this indirect evidence is more diagnostic of the physical water system than the few direct measurements a project budget can afford — because it captures the system at scales no targeted measurement can.
A few examples that any senior water professional recognizes:
Stream temperature reveals groundwater discharge
Cool, stable thermal regimes mark reaches fed by groundwater; warm, flashy regimes are surface-runoff dominated. A continental-scale analysis classified 1,729 US stream sites this way without running a physics model — using temperature records that already existed.
Cold-water fish confirm what the thermometer suggests
Trout, salmonids, and groundwater-dependent aquatic communities persist only where year-round cold groundwater discharges. The US Fish & Wildlife Service's species records are a hydrogeological dataset.
Phreatophyte vegetation marks the water table
Cottonwoods, willows, and riparian woodlands tap groundwater directly; satellite NDVI tracks their response to depth-to-water at landscape scale. Meinzer's 1927 USGS classic established plants as groundwater indicators — Sentinel-2 NDVI now operationalizes it everywhere.
Landform tells the subsurface story
Breaks-in-slope, depressions, oxbows, lineaments, and paleo-drainage signatures mark discharge zones, recharge corridors, fracture systems, and buried channels — patterns no head-field map will show as clearly.
Land use is a recharge signal
Forested catchments recharge differently than urban or irrigated landscapes; impervious fraction changes the runoff multiplier. The land-cover change history is the recharge change history.
Thermal infrared imagery finds discharge directly
Landsat 8 thermal bands and drone TIR identify cold-water springs feeding lakes, submarine groundwater discharge along coasts, and seasonal seeps — direct visual evidence of where groundwater reaches the surface.
Water chemistry fingerprints flow paths
Tritium, ³H/³He, ¹⁴C, CFCs, and stable isotopes date groundwater (years versus millennia), trace flow paths, identify mixing ratios, and assess aquifer vulnerability — chemistry the flow equations alone cannot recover.
InSAR subsidence is a pumping audit
Sentinel-1 has documented aquifer compaction across California's Central Valley, Mexico City, Tehran, and Lahore. Subsidence rates are direct measures of unsustainable extraction. The satellites were already there; the inference was free.
Archaeological patterns trace buried paleochannels
Kushan-era forts along Pakistan's Cholistan Desert mark the buried Hakra paleochannel precisely — the coarse-sediment recharge superhighway modern wells need. Civilizational settlement is a hydrogeological dataset collected at multi-millennial scale.
Historical records reveal what was
Well permits, irrigation rights, abandoned villages, and main-break records carry decades of hydraulic information no monitoring program collected systematically — but every one of which is publicly available.
None of the inferences above fit inside a physics equation; all of them refine the picture of how water actually moves and where it actually matters. The same networked data architecture then delivers the other half of the bridge to physics — the quantitative parameter and forcing fields the equations need.
Qualitative evidence conceptualizes the physics model
Diagnostic inferences shape what the model is even modeling — which processes matter, where boundaries belong, what kind of water system this is, where the system is sensitive. InSAR subsidence becomes the calibration target for compaction. Thermal-anomaly discharge zones become boundary conditions. Phreatophyte NDVI patterns inform where the simulated water table needs to match reality.
Quantitative layers parameterize the physics model
Dense controlling-data layers — terrain, soils, land cover, climate, hydrography, geology, infrastructure, live sensor feeds — are spatially distributed and equation-ready. They become the recharge, conductivity, boundary heads, evapotranspiration, channel networks, runoff coefficients, and observation series the physics platforms ingest as parameters, forcings, and calibration data.
Where qualitative evidence tells the modeler what kind of water system this is, the quantitative layers tell the equations how much water moves where. DataNET conceptualizes and quantifies the physics model before any equation is solved — qualitative evidence for what the system is, quantitative layers for what the equations need.
Each paradigm feeds the other.
The two paradigms feed each other. When process-based modeling's results are published as web services, they become another data layer in DataNET — fusable into the next project's data-driven model. The loop closes through published services. Process-based simulation also reveals which data gaps matter and which don't — so the user knows what additional information is worth collecting, and what is not. Not all data improves decisions; the loop helps users find the data that does.
Same data source, different cognitive object.
DataNET shows lithology data from well records as fused 3D cylinders — the data, styled. Each borehole becomes a stack of colored bands representing the lithologic units observed at that location, with depth, thickness, and unit type rendered visually. The user sees what the data says. IGW-NET goes further: it constructs a hydrogeologic model from the lithology data — assigning hydraulic properties (conductivity, storativity, porosity) to lithologic units, building the layered structure the groundwater equations require, then visualizing the constructed model in 3D. Same data source, different cognitive object: DataNET shows "this is what the borehole logs say"; IGW-NET shows "this is what the hydrogeologic interpretation looks like." Both are valid; they answer different questions for different audiences.
Two complementary routes to the modeling platforms. Essential data direct-linked; everything else, federated and on demand.
MAGNET4WATER's data infrastructure has two distinct routes into the four modeling platforms. They are complementary by design — and they are not the same substrate. The preprocessed multi-resolution database is direct-linked (no DataNET hop); DataNET is the federated route to the world's water-data services. Together they cover what each optimizes for.
Essential data, curated and pre-shaped, direct-linked into the modeling platforms.
The preprocessed multi-resolution database is HydroSimulatics' curated set of essential layer types — terrain (DEM), soils, land use, climate, hydrography, bedrock topography, hydraulic conductivity, recharge, water wells. Mostly global and continental coverage, with deeper national-US coverage for key layers, pre-shaped at the resolutions each physics platform needs (30m, 90m, 300m, 1000m and finer where available). IGW-NET, SwaNET, StormNET, and ConduitNET read this database directly. No DataNET hop in between. This is what makes a working model possible in minute ten: the platform-ready essentials are already pre-shaped to what each physics platform needs.
Everything else, live on the network.
DataNET federates worldwide water-data services — 145+ provider domains, 530+ service endpoints, 17K+ unique layers (73K+ layer records). Global services (USGS, NOAA, NASA, ESA, FAO, Copernicus). National services (national meteorological, geological, hydrological agencies). State, county, and local services. Corporate and private services published by individual organizations through GeoServer — accessible through DataNET when registered, optionally password-protected, and never required to appear on the public hub. HydroSimulatics' own GeoServer (the ~60,000 converted services described above) is one of the federated nodes. DataNET stores the URL catalog and metadata; the actual data stays with its owners, fetched live on demand. The federation hub is a smart catalog with a transfer engine, not a data warehouse.
Link by geometry, not by dataset.
The direct route is fast because it is pre-shaped to platform needs. DataNET is general because it links to platforms by geometry type, not by specific dataset. DataNET maps point data to platform point fields, line data to platform line fields, polygon data to platform polygon fields, raster data to platform raster fields — dynamically processed on transfer. If your IGW-NET subbasin needs a recharge raster, you can transfer any relevant raster from any federated service. If your SwaNET watershed needs a stream-network update, you can transfer any relevant line data. If your StormNET subcatchment needs an updated impervious-surface layer, the same. The user has to know what is relevant — the direct route gave the default; DataNET is the power-user path that brings any networked data into any compatible model field. A few extra keystrokes for the generality. Dynamic processing keeps high-resolution data usable even for large domains, dense grids, and complex cell systems.
Three uses for transferred data: as fabric — contribute to the platform's data context and visualization layers; as calibration — provide observations the model is tested against; or as input — provide model parameters, boundary conditions, forcing fields. Any data on the network that a model can use, can flow into that model. Any data → any model → any field.
The preprocessed database provides consistency and speed. DataNET provides flexibility, depth, and control.
Advanced users evaluate multiple sources — including imperfect, conflicting, indirect, or soft data. Not all data is clean. Not all data is consistent. But much of it is still useful — if it is interpreted correctly. DataNET is designed for that reality.
How data flows into your modeling work.
A user working in any MAGNET4WATER platform can draw on data from four distinct sources, in any combination.
1. Global preprocessed base
Built into each platform — DEM, soils, land use, hydrography, climate, bedrock topography — multi-resolution by construction. The direct route described above; available the moment you open any platform.
2. Curated DataNET hub
HydroSimulatics-harvested WMS/WFS/WCS services from USGS, NASA, NOAA, and many more — plus HydroSimulatics' own published services. Browse by region (global → continental → national → state → county), filter by keyword, grouped by category.
3. Direct upload through a platform
For project-specific data — a site survey, a custom DEM patch, a contaminant observation set, a private monitoring file. No server needed; the data lives with that model.
4. Your own GeoServer node
For organizational persistence, portfolio-scale consistency, and confidential data. Your services don't appear in DataNET's curated hub (it stays HydroSimulatics-managed), but you can point DataNET at your service URL — optionally password-protected — to visualize and analyze your data inside the same environment, and transfer it to any modeling platform. Your data stays on your own infrastructure; your server controls access. Fits naturally with confidential computing requirements.
Nine domains that drive water-system understanding.
Zoom into the federated Data Service Hub. Across nine domains, each pulls from many providers, scales, and standards — all reachable through one interface.
Terrain & Elevation
DEMs, LiDAR, bedrock topography. Foundational for watershed delineation, flood routing, drainage analysis, river networks, surface-groundwater interaction, hydraulic gradients, infrastructure siting, urban stormwater modeling.
Soils & Hydrologic Properties
One of the largest concentrations of model-ready hydrologic soil layers in the ecosystem — infiltration, recharge, runoff generation, erosion, vadose zone, agricultural water management, green infrastructure design.
Land Cover & Land Use
Drives hydrologic response, evapotranspiration, roughness, runoff. Supports urban growth analysis, flood studies, agricultural assessments, climate resilience, ecological analysis, stormwater planning.
Groundwater & Aquifer
Aquifer characterization, well analysis, recharge studies, saltwater intrusion, groundwater-surface water interaction, water supply planning, managed aquifer recharge. Especially central for IGW-NET workflows.
Climate & Weather
Meteorological forcing for watersheds, flood forecasting, recharge estimation, water supply, climate adaptation studies, drought assessment, environmental forecasting at local-to-global scales.
Geology & Lithology
Aquifer structure, confining units, fractured systems, sedimentary architecture, groundwater storage, subsurface heterogeneity, basin evolution. Essential for advanced hydrogeologic modeling.
Surface Water & Hydrography
River systems, lakes, reservoirs, stream networks, watershed boundaries, drainage connectivity, floodplain analysis — how water moves through connected natural systems.
Infrastructure & Built Environment
Urban planning, stormwater design, utility analysis, transportation interaction, water distribution systems, resilience planning. Increasingly important as natural and engineered systems become more interconnected.
Coastal & Ocean
Marine boundary datasets, coastal hazards, sea-level context, oceanographic forcing — for coastal infrastructure resilience, saltwater intrusion studies, and integrated land-sea analyses.
Every layer carries its full metadata — sources, abstracts, resolution, limitations, versions, publishers, references, search keywords — so the data is usable for serious work, not just visualization.
Display the world without a box. Process the site by drawing one.
DataNET operates simultaneously as a federated data service network (the foundation), an analytics and visualization environment (display tier), and a data-driven modeling system (processing tier). It has three entry points, organized by whether you are displaying data or processing it. Display does not require defining a site — the layers come at their native extent and the viewport is your window. Processing requires defining a site — the box tells DataNET over what area to transform data into something more than display. Two display modes; three site-focused outcomes from the box.
Browse the world's water-data services at their native extent.
For spatial exploration, layer comparison, and visual story-building — open the workspace. Layers arrive at the extent they were published at; the viewport is your window onto the world.
2D Map Workspace
The world's water-data services in one GIS-style 2D workspace. Pull layers from any federated service at their native extent — a global precipitation layer at global extent, a national soils layer at national extent, a county groundwater layer at county extent — and stack them as map overlays. Reorder, toggle, restyle, configure. The viewport is the window: pan and zoom across the whole world, or focus on a region of interest. The configuration of layers is the model — different question, different stack, different story.
Cesium 3D Globe — drape mode
Cesium's 3D globe view with WMS layers draped onto the terrain. Same federated layers as the Map Workspace, rendered in 3D globe context. Pan, tilt, fly, zoom — immersive globe-scale spatial intelligence. Virtual site visit before any field deployment. 3D GIS — map overlay on a globe. No processing happening yet; this is display only.
When you need to transform data, define a site.
Drawing a box defines the spatial extent over which DataNET processes data into something more than display. From there, three outcomes — three different ways to use the fused data. Two stay inside DataNET (Cesium styled features, VTK site model); one bridges to a physics platform.
Cesium styled-features mode
Box required. DataNET fetches federated WFS records inside the box and transforms them into 3D objects: wells become 3D cylinders with depth attributes; lithology becomes stacked colored bands along each well; monitoring stations protrude from the terrain with their attribute values; lines and polygons become 3D solids. WMS draped on the box terrain as the surface texture. The result is a globe-scale immersive site model — fused, styled, story-telling.
VTK Cartesian site model
Box required. DataNET fuses the site's subsurface data into a Cartesian 3D site model: DEM for the ground surface, bedrock surface as the lower boundary, additional elevation layers (intermediate stratigraphic surfaces, perched water tables), subsurface aquifer layers as the layered hydrogeologic structure, wells with screen intervals and completion details, water-quality observations at their sample locations. WMS imagery draped as DEM textures. The site-scale subsurface story, fused from federated data.
DataNET visualizes data. It does not construct interpretive models. Iso-surfaces, 3D plume concentration surfaces, 3D flow vectors, and model-structure 3D lithologies belong to IGW-NET, where lithology becomes hydraulic-property layers in a flow simulation.
Transfer to physics platform
Box required. Map federated layers to platform fields by geometry type — point to point, line to line, polygon to polygon, raster to raster. Dynamically process on transfer. Send the data into IGW-NET, SwaNET, StormNET, or ConduitNET as fabric, calibration, or input.
More data does not automatically create better understanding.
The value comes from choosing what to include, what to exclude, how to render it, and what story the fused evidence supports. Fusion is a design problem — curation, not accumulation.
Curate, don't stack
Effective characterization is intentional. Layers are selected, ordered, styled, and accompanied by metadata that explains what each adds. More is not better; better is better.
Reveal relationships
Fusion exposes patterns: how terrain shapes hydrography, how soils control infiltration, how monitoring data confirms or contradicts model assumptions, how vulnerability maps emerge from layered evidence.
Communicate insight
Map Workspace and Cesium views are not just for analysts — they are for decision-makers, stakeholders, and the public. The same fused workspace serves analysis and communication.
Connect maps to monitoring networks.
DataNET connects to national telemetry and sensor networks: USGS National Water Information System (NWIS) live-linked via direct API in the US, Government of Canada water sensor networks live-linked via direct API in Canada (parallel architecture) — groundwater monitoring wells, stream gages, water-quality stations, realtime feeds, historical records, and time-series services. Telemetry data overlays the same workspace for local and regional statistical analysis, time-series visualization, and instant analytics — the data-driven companion to model calibration in IGW-NET and SwaNET.
From library discovery to model-ready inputs.
DataNET supports the full data lifecycle: finding layers across the federated network, building workspaces, preparing 3D visualizations, transferring selected data to modeling platforms, and using the results in Model Observatory reports and decisions.
Find
Browse the hierarchical Data Tree by region, category, keyword, and metadata. Search across federated agencies and HydroSimulatics-published services.
Organize
Pull selected layers into the Map Workspace at their native extent. Stack, reorder, restyle, configure. Save workspaces for reuse across sites or programs.
Visualize
Display the workspace as 2D map, drape it on the Cesium globe, or draw a box to process the site as Cesium styled features or VTK Cartesian site model.
Transfer
Send selected layers to IGW-NET, SwaNET, StormNET, or ConduitNET — by geometry type — as fabric, calibration, or input. (See Section 6 for the full cross-platform handoff.)
Report
Use fused data and model results in Model Observatory reports — closing the loop between data, analysis, and decision.
The federated route into IGW-NET, SwaNET, StormNET, and ConduitNET.
DataNET is one of five MAGNET4WATER platforms, and it is also the federated data route serving the other four — alongside the always-on preprocessed multi-resolution database that direct-links to all four modeling platforms separately from DataNET. DataNET supplies federated WMS/WFS/WCS services, observations, and the any-data-to-any-model-field transfer engine.
Four steps from draw-a-box to platform input.
Draw an analysis area
Define a rectangle or polygon for the site, basin, city block, watershed, or project area.
Select relevant data
Add layers from the Data Tree or workspace: WMS images, WFS vectors, WCS rasters, local uploads, or screenshots.
Transfer to modeling platforms
Send DEMs, land use, soils, rivers, wells, images, boundaries, or reference layers to the platform where they become modeling context or input.
Refine through feedback
Return to DataNET to add layers, adjust context, improve conceptual understanding, and refine the next modeling pass.
IGW-NET — Federated groundwater context
Through DataNET, IGW-NET can transfer in: agency well networks (USGS NWIS, Government of Canada monitoring-well networks), recent recharge and land-use scenarios from regional providers, project-specific bedrock topography, water-quality observations from state agency portals.
Open IGW-NET →SwaNET — Federated watershed context
Through DataNET, SwaNET can transfer in: PRISM 800m precipitation, NCEI Access API rain-gauge time series (coming next release), regional climate downscalings, project-specific soils and land-use overrides, monitoring stations from agency networks.
Open SwaNET →StormNET — Federated urban context
Through DataNET (coming next release): municipal GIS layers, infrastructure inventories from agency portals, NOAA Atlas 14 IDF curves (already directly linked), regional IDFs (GSDR-IDF, ECCC IDF) coming next release, recent flood mapping.
Open StormNET →ConduitNET — Federated distribution context
Through DataNET (coming next release): service-area boundaries, demand pattern datasets from regional providers, operational layers from agency portals.
Open ConduitNET →Stop rebuilding data infrastructure. Start understanding water systems.
Stop rebuilding what is already there. The world's water data is networked, the modeling platforms are built, the bridge between them is DataNET. Three "Better" outcomes follow when teams stop fragmenting the workflow into per-project silos. The bigger shift: water-systems understanding meets professionals in the working language they already use — planning, regulation, utility operations, hydrogeology, ecology, infrastructure management — at whatever technical depth the question requires.
Better questions
Users see patterns and gaps earlier, leading to better hypotheses and modeling priorities.
Faster insight
Virtual site visits and data fusion identify major controls, vulnerabilities, and constraints before simulation.
Better models
Models built after spatial characterization are better grounded: clearer boundaries, better inputs, fewer blind assumptions.
Understand the system first. Model it second.
DataNET is the entry layer, intelligence layer, and integration layer of MAGNET4WATER.
From classroom to consulting to management.
Education
Bring big data exploration into classrooms routinely.
Consulting
Reduce site characterization and conceptual model development costs.
Management
Help decision-makers see relationships without waiting for a full model.