The Flow has not summarised this story yet — read it at arXiv Machine Learning.
The paper introduces a workflow that ranks competing geological interpretations by converting them into spatial priors and assessing their consistency with hydraulic‑head observations. Using a text‑to‑image model to generate 1600 geologic images per interpretation, a variational autoencoder encodes them, and a supervised inverse network maps head data into this latent space, producing log‑conductivity fields for flow simulation. The method is validated on a synthetic Johansen Formation benchmark and applied to two conceptual models of the Culebra Dolomite, yielding compatibility scores that align with independent evidence.
arXiv:2605. 03383v2 Announce Type: replace Abstract: Geological interpretation infers subsurface properties and structures from indirect geophysical observations.
arXiv:2609.36193v1 Announce Type: new Abstract: Learning from scientific measurements often requires aligning modalities with different spatial support and resolution. Subsurface characterization is...
arXiv:2606. 24967v1 Announce Type: new Abstract: In ill-posed inverse problems, the recovered solution depends as much on the prior as on the data, yet much of the engineering knowledge that could serve as that prior is recorded qualitatively rather than in formal mathematical form.
arXiv:2607. 16050v1 Announce Type: new Abstract: Pluvial (rainfall-driven) flooding accounts for 45% of National Flood Insurance Program (NFIP) claims in the United States and is harder to predict than its riverine and coastal counterparts, with existing approaches limited to coarse resolution, regional domains, or computationally intensive process-based models unsuitable for daily continental-scale use.
arXiv:2603.25779v2 Announce Type: replace-cross Abstract: Groundwater represents a key element of the water cycle, yet it exhibits complex and context-dependent relationships that make its modeling c...