Excited to share new paper on Nature Reviews @NatRevEarthEnv. Want to discover relationships from data? Or answer questions in #Geosci? Or obtain ML-level performance with interpretable processes? Differentiable Modelling could be your answer! (1/n)

At @iclr_conf the AI conf today. Anyone in Singapore? My presentation is tmr.

RT @KSawadekar: Excited to share our new paper in Journal of Hydrology! .We developed a differentiable model that fuses multiple precipitat….

A massive paper and a generational leap in national-scale high-resolution seamless hydrologic simulation. We published the continental-scale simulation dataset. If you have interests, please inquire! ./run

Dear all, we are extending the abstract submission deadline to April 1. After this date, we will evaluate the situation. Therefore, we kindly ask that you refrain from booking travel until you hear from us regarding the next steps.

RT @kcc_carroll: New paper alert "Hydrogeological.Modeling With an Integrated Latent Diffusion Framework" with @reza_soltanian, Chuanjun Zh….

RT @danrothenberg: From. The. Rooftops. You're not going to find a magic money-printing machine anywhere else that will give you these ty….

meanwhile please follow me on Linked Internally and Ciel Bleu.

Dear all,.Since many are here, the news is here. HydroML 2025 will be hosted at UC Irvine and held at the UCLA Lake Arrowhead Lodge near Los Angeles. It is the week of May 26 - 30 (likely May 27 - 29) Final dates TBD. Check out

RT @SteveBurianH2O: Great opportunity with CIROH.👇👇👇.

Please come and evaluate our continental-scale seamless streamflow simulation (1990-1999 more is being uploaded) using hybrid physics-informed ML ("differentiable") hydrologic model. Could be a blast!.This comes with our preprint.

@Ambehroozi meanwhile, please follow me where sky is bluer. I'm gonna be on both sides through the transition.

This is our first submission to a generic AI conference. Can't wait to bring our unique perspectives into the main AI realm. @Ambehroozi is the first author of the paper. Stay tuned for the paper --- will make some minimum edits for the camera-ready version!. #ICLR2025.

Just post some news until other platfms are setup. Our paper  “Sensitivity-Constrained Fourier Neural Operators for Forward and Inverse Problems” has been accepted into ICLR 2025 main, a premier AI conference. It is so hard and rare for domain people to publish here.

Examples and tutorials are given for doing differentiable hydrologic modeling, soon ecosystem modeling, water temperature modeling, and more… this is the framework to build on! Ongoing work will add data assimilation, multiscale modeling, unique surrogate models, etc….

dMG is not a model. Rather, it is a framework to uniformly support differentiable models (training neural networks together with process equations together) while integrating an ecosystem of services. Should be very easy learning curve. Pick up differentiable modeling in hours!.

Don’t know why but whenever I type in these text boxes on my phone there can be lots of typos and grammar bugs that i failed to detect….darn it… sorry folks, but the dMG framework is one true gem!.

We are release the Differentiable Modeling - Generic (δMG) framework, a domain-agnostic, PyTorch-based framework for developing models that can train neural networks with process-based equations together on big data. It unifies many DM research activities.

Have been posting minimum here. This is our group's AGU presentations. Hope to see you. I've a talk on Monday and one on Friday. Also a poster on Monday. Also, we are gearing up some a major code release this week. Drumming.