iiinova — where your digital self keeps evolving

Chaopeng Shen

@chaopengshen.bsky.social

Prof @PennStateCEE. Solving water and climate resilience problems with AI & physics-informed AI. Rivers, soil moisture, groundwater, landslide, all connected. Rethink H2O. Mhpi.info

Posts Replies Media

Chaopeng Shen @chaopengshen.bsky.social · 27/01/2025

Please come and evaluate our continental-scale seamless streamflow simulation (1990-1999 more is being uploaded) using hybrid physics-informed machine learning ("differentiable") hydrologic model. Could be a blast! zenodo.org/records/1429... This comes with our preprint. doi.org/10.22541/ess...

zenodo.org

Streamflow datasets from the high-resolution, multiscale, differentiable HBV hydrologic models

This dataset is from the high-resolution, multiscale, differentiable HBV hydrologic models, provided by the Multi-scale Hydrology, Processes, and Intelligence (MHPI) team from The Pennsylvania State U...

020

Chaopeng Shen @chaopengshen.bsky.social · 27/01/2025

Not surprising some new forms of AI requires much less energy to train. Basically I think the form of the AI has not settled yet. There could be way more energy-efficient forms of AI down the road. Think about it. How much energy does a person's brain take to logically think and answer a question?

000

Chaopeng Shen @chaopengshen.bsky.social · 26/01/2025

doi.org/10.1029/2023... CUAHSI summer institute product: A random-forest model to predict river geometry (width and depth) of all river reaches in USA. Sagy Cohen and I advised the students (from other research groups) in summer 2022 over several weeks and the paper is out in 2024.

doi.org

The Geometry of Flow: Advancing Predictions of River Geometry With Multi‐Model Machine Learning

Machine Learning models outperform regional (physiographic) hydraulic geometry equations for predicting stream width and depth Model performance varies by stream orders and geographical regions, ...

010

Chaopeng Shen @chaopengshen.bsky.social · 22/01/2025

I won't come here empty-handed. Our paper “Sensitivity-Constrained Fourier Neural Operators for Forward and Inverse Problems” has been accepted into ICLR 2025 main, a premier AI conference. Stay tuned for the paper --- will make some minimum edits for the camera-ready version!

122

Chaopeng Shen @chaopengshen.bsky.social · 22/01/2025

My first post on Bluesky! Not easy for me to face a new platform and start from scratch, so any help from friends are appreciated!!

030