Environmental & Water Researchers

The platform was built by academic researchers, for academic researchers — wrapping the federal-agency engines the community already uses and trusts. MODFLOW 6 (USGS). EPA SWMM. EPA EPANET. USDA SWAT. MT3DMS. MODPATH. SEAWAT. T-PROGS. UCODE. The same code your most cited papers run on, in a cloud-native environment where reproducibility, multi-scale coupling, and open-science publication are first-class concerns — not retrofits.

The 2026 water-research pressure pattern
Open-science mandate

The OSTP August 2022 memo requires federally-funded research to publish results and data openly. NSF, NIH, DOE data-management plans now demand archived code and reproducible workflows. The unit of deliverable is no longer "a paper" — it is "a paper plus a reproducible model plus a reproducible dataset."

Reproducibility crisis

A 2023 Nature survey found roughly 70% of researchers had failed to reproduce another scientist's results, with reproducibility now treated as a structural credibility problem for the field. Hydrology has not been immune. Methodology that cannot be re-run is methodology that cannot be cited with confidence.

Code-repository requirements

Nature, Science, PNAS, Water Resources Research, Hydrology and Earth System Sciences now require archived code and data alongside the manuscript. Zenodo, Figshare, HydroShare DOIs are routine. The model file IS the supplementary material — not a description of one.

AI/ML in hydrology

Foundation models in earth science (Pangu, ECMWF AIFS, NASA Prithvi, ClimSim) are reshaping the field. NSF program announcements increasingly ask how proposals integrate AI/ML with process-based modeling. The methodological question is no longer "ML or physics" but "how do they couple."

Federated collaboration

Big-team water-science projects span institutions, continents, and time zones. EU Destination Earth, NSF Big Data Earth Science programs, NOAA-supported CUAHSI/CZNet infrastructure, multi-country Horizon Europe consortia. The unit of work is no longer a single lab but a federated team.

Tenure beyond papers

Research universities increasingly count software, datasets, dashboards, and public-facing observatories as first-class research products alongside peer-reviewed papers. The deliverable that earns a citation is also the deliverable that builds a case for tenure or NSF CAREER award.

These pressures land on the same lab in the same year: a PhD student writing a thesis, a postdoc writing a paper, a PI writing a grant. Every analytic infrastructure decision touches all three timelines.

What Sets MAGNET Apart

Four architectural decisions that map directly to how academic water research is actually done.

⚙️

Community engines. Underneath. Throughout.

USGS MODFLOW 6 for 3D groundwater flow. MT3DMS for reactive contaminant transport. MODPATH for particle tracking. SEAWAT for variable-density flow. T-PROGS for transition-probability geostatistics. UCODE for inverse-problem calibration. EPA SWMM for stormwater and combined sewer. EPA EPANET for distribution networks. USDA SWAT for watershed modeling. The platform wraps the federal-agency engines the community already cites — without altering the underlying physics. Reviewer 2 can verify the methods.

🪆

Multi-scale, multi-resolution — properly coupled.

"Models in a model" is a published methodological problem in nested modeling, downscaling, and multi-scale coupling. MAGNET's hierarchical architecture supports parent-child model nesting with documented handoffs of physically-meaningful quantities — head, flux, concentration, recharge — at the interface. Run a regional watershed model and a site-scale aquifer model in the same study, with the boundary conditions actually flowing between them rather than being asserted separately.

🔁

Reproducibility built in.

Every model is version-controlled with documented data lineage from federated sources (USGS NWIS, NASA Earthdata, NOAA, state monitoring networks) through preprocessing through simulation through visualization. Publish to a public Observatory page with a citable DOI; archive the model file alongside the paper. What you write in your methods section can be re-run by a reviewer from your supplementary materials — not described, re-run.

🎓

Free academic tier.

Faculty, postdocs, graduate students, and undergraduates at accredited institutions can use the platform free for instruction and non-commercial research. Lower the barrier to graduate-student adoption; remove the budget conversation from grant-supported research; let labs build research workflows around capabilities rather than licensing constraints. HydroSimulatics is a Michigan State University spin-off — the academic relationship is structural, not transactional.

Right physics for the question

Water research spans physically distinct domains. Each MAGNET platform handles its domain on the community-standard engine for that physics — coupled through documented, physically-meaningful interface variables that can survive cross-examination by a reviewer.

Groundwater & transport
IGW-NET
3D flow, contaminant transport, particle tracking, variable-density, heterogeneous geology, inverse-problem calibration — USGS MODFLOW 6 + MT3DMS + MODPATH + SEAWAT + T-PROGS + UCODE
Watershed & climate
SwaNET
Process-based hydrology, sediment, nutrients, climate downscaling (CMIP6), land-use change, BMP analysis — USDA SWAT
Urban hydrology & CSO
StormNET
1D unsteady Saint-Venant, combined sewer, LID with multi-layer physics, flood inundation, buildup-and-washoff — EPA SWMM, FEMA-approved
Distribution networks
ConduitNET
Pressurized-flow networks, water-quality decay, demand pattern analysis, energy minimization — EPA EPANET
Federated data fabric
DataNET
USGS NWIS, NASA Earthdata, NOAA, state monitoring networks, CMIP6, federated WMS/WFS/WCS — the data ecosystem your peer reviewers also use
What MAGNET is — and is not

MAGNET4WATER is an analytic infrastructure that wraps community-standard engines in a cloud-native, data-centric, reproducible environment. It does not replace methodological rigor, peer review, the scientific judgment of the investigator, or the difficult interpretive work of relating model output to the underlying physical system. What it adds is the data-engineering, multi-scale-coupling, visualization, and publication infrastructure that water research has historically had to build separately for every project. The hypothesis testing, model formulation, and interpretation remain where they belong: with the researcher.

Pain Points & MAGNET Solutions

1

Multi-scale coupling methodology

Coupling a regional-scale watershed model to a site-scale aquifer model — with boundary conditions that actually flow between them rather than being asserted separately — is a hard methodological problem that consumes substantial PhD-thesis time per study

MAGNET: Hierarchical parent-child model nesting with documented handoffs of physically-meaningful interface variables. SwaNET watershed model passes recharge to IGW-NET aquifer model; aquifer baseflow returns to surface-water network; sediment and nutrients flow consistently across scales. The "models in a model" architecture solves a published methodological problem — not a marketing capability.

2

Reproducibility & open-science compliance

NSF / NIH / DOE data management plans now require archived code and reproducible workflows. High-impact journals require Zenodo or HydroShare DOIs alongside manuscripts. Most lab workflows weren't built for this

MAGNET: Every model is version-controlled with documented data lineage from federated sources through preprocessing, simulation, and visualization. Publish to an Observatory page with citable DOI; archive the model file alongside the paper. The reproducibility is structural, not a compliance overlay applied at the end of the project.

3

Data-engineering tax

A graduate student's first year of a hydrology PhD often consists of downloading datasets, re-projecting CRS, debugging GIS pipelines, and parsing inconsistent file formats — before any science begins

MAGNET: Preprocessed hierarchical global base model with federated WMS/WFS/WCS access. The data ecosystem — USGS NWIS, NASA Earthdata, NOAA, CMIP6, state monitoring — is already live-linked and integrated. The data-engineering tax goes from the first year of a PhD to the first afternoon. Grad students spend time on science, not format wrangling.

4

Cross-institutional collaboration

Big-team water-science projects span institutions, continents, time zones, and computational environments. Setting up shared analytic infrastructure is its own multi-month sub-project

MAGNET: Browser-based; no installation required at any institution. Shared models in the Observatory with role-based access. Confidential modes for pre-publication work; public modes for shared open-science output. A Helmholtz lab, a CGIAR center, and an NSF-funded R1 university work on the same model in the same environment — without negotiating computational stacks at each site.

5

Methodology consistency across grant lifetimes

A typical NSF grant runs 3 years; CAREER 5 years; large center grants 10. Over that span, postdocs cycle, PhD students graduate, codebases drift, and the methodology described in Year 1 isn't always the methodology in the Year 5 final report

MAGNET: Federal-agency engines underneath remain stable across the grant lifetime. Configuration and methodology are version-controlled in the platform; institutional knowledge gets codified in the platform setup, not in retiring postdocs' notebooks. The final report describes the same methodology the initial proposal did, on the engines federal science has maintained for decades.

6

Communicating to non-experts

An impact statement for a press office. A public hearing testimony. A briefing for a Congressional staffer. The same model that produces journal figures has to produce something a non-expert audience can interpret — without becoming a different model

MAGNET: 3D streaming visualization, time-evolving animations, and AI-generated explanatory reports grounded in the actual model file. Publish to a public Observatory page. The visualization that lands in the press release is the same one that produced the journal figure — with the underlying analysis defensible at both audiences.

Strategic Value

For faculty principal investigators, postdocs, graduate students, and the research community as a whole, the architectural commitments translate into three dimensions: research velocity — more hypotheses tested per FTE-year, with graduate students spending time on science rather than data engineering; research products that count toward tenure — code, datasets, dashboards, and citable observatories as first-class outputs alongside peer-reviewed papers; and methodological credibility — community-standard engines underneath, reproducibility structural rather than retrofitted, multi-scale coupling on documented physically-meaningful interface variables.

HydroSimulatics Inc. is a Michigan State University spin-off; the founding team is faculty. The platform was built inside the academic research community before it was built for it. That is the credential vendor marketing cannot fake.

Proven for Research and Teaching

Three deployments demonstrating the architectural patterns water research and teaching actually run on: undergraduate-to-research bridge, applied-science portfolio methodology, and the MSU institutional partnership the platform was built within.

Launch Platform — Free Academic Tier Consulting Research Partnership See Pricing
For faculty PIs and research-group leaders

Why does the multi-tier, multi-scale framework matter to a hydrologist or environmental scientist? Because it is the published research problem the field has been writing papers about for two decades — from regional-scale impact assessment to site-scale process modeling, from individual investigator-driven research to federated team science. Read the strategic argument for why this architectural moment is now — and how it changes the unit of analysis from "a model" to "a model network."

Read: The Inflection Point →
← Sectors↑ TopCase Studies →