Our work

Projects & Research

Explore EnviTrace’s documented research and development programs, applied studies, and scientific foundations. Each entry connects a short summary with a representative image and supporting sources.

8 of 8 entries

Browse entries by name

Water & environmentGenesis Mission research

RAIN

EnviTrace leads RAIN, the Rigorous AI and Information Network for Surface-Subsurface Water System Coupling, under the U.S. Department of Energy Genesis Mission.
The research connects monitoring observations, physical models, scientific computing, and AI to study how surface water and groundwater interact.
Its public dashboard brings together ten Colorado River watersheds, station observations, weather, snow, groundwater, and reservoir histories.
Saved streamflow comparisons make model behavior inspectable while retaining the distinctions between training, validation, testing, and retrospective evidence.

Geothermal energyGenesis Mission Phase II research

MAESTRO

EnviTrace is an industry partner in this UC Irvine-led, three-year Phase II Genesis Mission project for enhanced geothermal systems.
MAESTRO will coordinate specialist AI agents, field observations, and physics-based simulations to reason about complex subsurface conditions.
The planned research addresses fractures, reservoir connectivity, coupled physical processes, uncertainty, and induced seismicity.
Planned tools include EGSChat for conversational interaction with reservoir analyses and a browser-based GIS copilot for site screening.

Seismicity & subsurface riskDOE-supported development

LAPIS

The DOE-supported LAPIS program develops physics-informed AI workflows for subsurface characterization and induced-seismicity research.
The work combines seismic observations, well and injection information, and physical context to examine how subsurface activity relates to seismic response.
SeisDB brings seismic and well information into a shared data workflow, supporting reproducible analysis and model development.
EnviTrace’s published updates describe collaboration with Lawrence Livermore National Laboratory and continuing research, testing, and opportunities for early adopters.

Geothermal energyDOE Phase II development

GeoTGo

EnviTrace announced DOE Geothermal Technologies Office Phase II support to advance geothermal exploration and development workflows built on its Phase I work.
The program integrates geological, geophysical, and geochemical information with physics-informed AI to investigate geothermal prospectivity.
Research priorities include reservoir characterization, drilling-target assessment, and transparent comparison of resource-development options.
Environmental and energy justice, open methods, and collaboration with practitioners are stated priorities of the development effort.

EnviCloud three-dimensional geological modeling demonstration showing synthetic wells, geological horizons, and a fault — Representative EnviCloud 3D modeling demo with synthetic geology; this illustrates the research area rather than a field-validation result.

Subsurface resourcesDOE award announced in 2024

GeoML: Geologic reservoir AI research

EnviTrace announced a $1.1 million DOE Office of Science award to develop AI software for modeling and managing geologic reservoirs.
The published project aims focus on integrating data, advancing physics-informed AI, and improving characterization and prediction.
Decision-making under uncertainty is a central research objective, connecting subsurface understanding with resource and operational choices.
Related GeoML work addresses multiple reservoir uses, including energy extraction and storage, carbon sequestration, and groundwater resources.

Urban energyPrototype research

UrbanAI

UrbanAI investigates how AI can connect urban environmental and socioeconomic data with questions about energy production and use.
Research themes include renewable energy integration, geothermal heating and cooling, urban heat islands, and local climate conditions.
The published roadmap describes collaboration with national-laboratory and university researchers and a prototype for preliminary testing.
Scientific plots expose data relationships and extracted features; real-time assimilation and wider deployment remain development goals.

EnviTrace Thermal Model: Published EnviTrace thermal-model dashboard in dark mode, with geographic model layers. — Published EnviTrace thermal-model dashboard in dark mode, with geographic model layers.

Applied research studiesOngoing research prototype

EnviTrace U.S. Thermal & Stress Model

This study explores geothermal temperature across the conterminous United States and through depth using a physical baseline plus a learned correction.
A baseline geothermal field represents broad physical structure before machine learning is applied to differences between that field and measured temperatures.
The interface allows users to inspect spatial predictions and vertical temperature profiles alongside the observations used to inform the model.
The published work is ongoing R&D, offering a way to inspect baseline, residual, and combined predictions rather than a finished resource assessment.

Published maps grouping Colorado River Basin subwatersheds by dry-day patterns across historical and future climate scenarios — Figure 6, reproduced unchanged from Talsma, Bennett & Vesselinov (2022), Earth and Space Science. © 2022 The Authors; CC BY 4.0. Colors identify groups within each panel; blank panels mark cases without an acceptable NMFk solution.

Research foundationsPublished study · 2022

Colorado River Basin drought patterns

This peer-reviewed study co-authored by EnviTrace CTO Velimir Vesselinov is part of the scientific foundation presented on the RAIN project page.
Researchers combined climate projections, a process-based hydrologic model, and NMFk to examine drought behavior across 134 Colorado River Basin subwatersheds.
Extracted spatial groups and seasonal signals show how dry-day patterns vary between historical and future periods under different climate scenarios.
The findings connect climate scenarios with interpretable drought patterns and provide a foundation for RAIN’s investigation of coupled water systems.

Loading…

Pipelining text and images

RAIN

EnviTrace leads RAIN, the Rigorous AI and Information Network for Surface-Subsurface Water System Coupling, under the U.S. Department of Energy Genesis Mission.

The research connects monitoring observations, physical models, scientific computing, and AI to study how surface water and groundwater interact.

Its public dashboard brings together ten Colorado River watersheds, station observations, weather, snow, groundwater, and reservoir histories.

Saved streamflow comparisons make model behavior inspectable while retaining the distinctions between training, validation, testing, and retrospective evidence.

MAESTRO

EnviTrace is an industry partner in this UC Irvine-led, three-year Phase II Genesis Mission project for enhanced geothermal systems.

MAESTRO will coordinate specialist AI agents, field observations, and physics-based simulations to reason about complex subsurface conditions.

The planned research addresses fractures, reservoir connectivity, coupled physical processes, uncertainty, and induced seismicity.

Planned tools include EGSChat for conversational interaction with reservoir analyses and a browser-based GIS copilot for site screening.

LAPIS

The DOE-supported LAPIS program develops physics-informed AI workflows for subsurface characterization and induced-seismicity research.

The work combines seismic observations, well and injection information, and physical context to examine how subsurface activity relates to seismic response.

SeisDB brings seismic and well information into a shared data workflow, supporting reproducible analysis and model development.

EnviTrace’s published updates describe collaboration with Lawrence Livermore National Laboratory and continuing research, testing, and opportunities for early adopters.

GeoTGo

EnviTrace announced DOE Geothermal Technologies Office Phase II support to advance geothermal exploration and development workflows built on its Phase I work.

The program integrates geological, geophysical, and geochemical information with physics-informed AI to investigate geothermal prospectivity.

Research priorities include reservoir characterization, drilling-target assessment, and transparent comparison of resource-development options.

Environmental and energy justice, open methods, and collaboration with practitioners are stated priorities of the development effort.

GeoML: Geologic reservoir AI research

EnviTrace announced a $1.1 million DOE Office of Science award to develop AI software for modeling and managing geologic reservoirs.

The published project aims focus on integrating data, advancing physics-informed AI, and improving characterization and prediction.

Decision-making under uncertainty is a central research objective, connecting subsurface understanding with resource and operational choices.

Related GeoML work addresses multiple reservoir uses, including energy extraction and storage, carbon sequestration, and groundwater resources.

UrbanAI

UrbanAI investigates how AI can connect urban environmental and socioeconomic data with questions about energy production and use.

Research themes include renewable energy integration, geothermal heating and cooling, urban heat islands, and local climate conditions.

The published roadmap describes collaboration with national-laboratory and university researchers and a prototype for preliminary testing.

Scientific plots expose data relationships and extracted features; real-time assimilation and wider deployment remain development goals.

EnviTrace U.S. Thermal & Stress Model

This study explores geothermal temperature across the conterminous United States and through depth using a physical baseline plus a learned correction.

A baseline geothermal field represents broad physical structure before machine learning is applied to differences between that field and measured temperatures.

The interface allows users to inspect spatial predictions and vertical temperature profiles alongside the observations used to inform the model.

The published work is ongoing R&D, offering a way to inspect baseline, residual, and combined predictions rather than a finished resource assessment.

Colorado River Basin drought patterns

This peer-reviewed study co-authored by EnviTrace CTO Velimir Vesselinov is part of the scientific foundation presented on the RAIN project page.

Researchers combined climate projections, a process-based hydrologic model, and NMFk to examine drought behavior across 134 Colorado River Basin subwatersheds.

Extracted spatial groups and seasonal signals show how dry-day patterns vary between historical and future periods under different climate scenarios.

The findings connect climate scenarios with interpretable drought patterns and provide a foundation for RAIN’s investigation of coupled water systems.