Interactive environmental analytics demos

Science-informed ML model optimizing energy production and storage.

Science-informed ML model optimizing carbon capture and storage strategies.

Geysers geothermal field data related to induced seismicity. Geysers is the largest geothermal field in the world.

Data imputation of the Great Basin geothermal datasets.

Use a bounded RobustML neighbor model to fill missing measurements along a geological well log.

Use bounded distance-weighted RobustML neighbors to fill gaps in a compact geological field sample.

Geothermal ML model inputs and results for continental U.S.

Geo modeling workflow demo.

Nonnegative Matrix Factorization (NMFk) for feature extraction and clustering. Extraction of hidden temporal (1D) signals.

Nonnegative Matrix Factorization (NMFk) for feature extraction and clustering. Exploration of spatial factors, clusters, and metrics.

Compare tensor resolutions using information retention, structure, and computational cost.

See how patterns and useful detail appear in data along a line and across a surface.

Explore a synthetic mine-scale lithium ranking through drill evidence, support-aware maps, cross sections, and spatial holdout checks.

Calibrate a model, check held-back data, and reveal influential or unidentifiable inputs.

Infer a release window and groundwater transport speed from monitoring-well histories.

Use NMFk to separate mixed groundwater chemistry into hidden source fingerprints.

Physics-informed forecasting and optimization for a two-well geothermal doublet. Set an operating priority and see the recommended strategy, its energy gain and its long-term thermal cost.

Combine Bayesian evidence with information-gap stress testing to compare contaminant source-termination choices under deep uncertainty.

Explore cost, carbon, and freshwater tradeoffs in a synthetic water–energy–food planning scenario.

Replay seeded offline-emulator results and audit uncertainty, evidence budgets, and abstention decisions for quantum ML workflows.