GeoBrain-Project

GeoBrain-Project / GeoBrain

Public

An End-to-End Differentiable Platform for Integrated Subsurface Modeling

geobrain-project.github.ioGeoBrain deep-learning differentiable-programming rock-physics seismic-inversion subsurface-modeling
21
4
100% credibility
Found Feb 25, 2026 at 20 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

GeoBrain is an open-source Python platform for geoscientific subsurface modeling that integrates differentiable physics simulations, Bayesian inference, and deep learning for characterization and inversion workflows.

How It Works

1
🔍 Discover GeoBrain

You hear about GeoBrain, a friendly tool that helps everyday people model what's underground like rock layers and fluids without needing to be a coding expert.

2
📦 Set it up quickly

Download and install it with simple steps, and soon your computer is ready to explore subsurface worlds.

3
📊 Add your measurements

Upload well logs or seismic recordings from your site, and GeoBrain organizes them neatly for you.

4
🪨 Build your earth model

Piece together rock types, fluid mixes, and layer shapes using guided examples that feel like building with blocks.

5
🌊 Watch physics come alive

Run wave or flow simulations and see ripples or fluids move through your model in real time – it's mesmerizing!

6
🔧 Tune to match reality

Adjust the model so its predictions perfectly fit your real data, with smart hints guiding you.

7
📈 View stunning results

Generate clear maps, animations, and uncertainty clouds showing exactly what's below ground.

✅ Make confident decisions

Now you have a reliable picture of the subsurface for drilling, storage, or exploration – ready to act!

Sign up to see the full architecture

6 more

Sign Up Free

Star Growth

See how this repo grew from 20 to 21 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is GeoBrain?

GeoBrain is a Python/PyTorch platform for end-to-end differentiable subsurface modeling, letting you chain geomodeling, rock physics, wave propagation, and reservoir flow into fully differentiable workflows. It handles everything from generating realistic geological fields to seismic inversion and Bayesian uncertainty quantification, solving the pain of stitching together siloed geoscience tools. Users get plug-and-play physics simulations that backpropagate gradients through the entire pipeline for optimization.

Why is it gaining traction?

Its standout hook is true end-to-end differentiability across multiphysics—from geostatistics to full waveform inversion—without manual gradients, plus 70+ rock physics models and samplers like SVGD/HMC for rigorous UQ. Real-world examples for CO2 sites (Illinois Basin, Sleipner) show immediate applicability, and the modular design means you add custom physics without rewriting code. Early adopters praise the 15 tutorials bridging scripts and notebooks for quick prototyping.

Who should use this?

Geophysicists tackling seismic/resistivity inversion, reservoir engineers running history matching or VOI analysis, and researchers in differentiable physics for subsurface apps. Ideal for those modeling CO2 storage or joint inversions where uncertainty matters, especially if you're tired of brittle Fortran simulators or non-differentiable Python stacks.

Verdict

Promising alpha-stage tool for niche end-to-end subsurface workflows, but with only 19 stars and 1.0% credibility, it's raw—expect bugs and sparse tests. Grab it if you're in geophysics prototyping; otherwise, watch for maturity.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.