Differentiable neural-network solver for data assimilation of ice shelves written in JAX

For downloading and calibrating IceSAT-2 and Bedmap thickness data

repository for work related to identification of streams and fractures on the Greenland Ice Sheet surface

Hello! This the is toy project for new comers uploading their first change.

Python library to train PINNs for 1D ice shelf modeling. Used in the paper "1D Ice Shelf Hardness Inversion: Clustering Behavior and Collocation Resampling in Physics-Informed Neural Networks" by Iwasaki and Lai.

Codes associated with the manuscript titled "Multi-stage neural networks: Function approximator of machine precision"

Accompanying code for paper: "1D Ice Shelf Hardness Inversion: Clustering Behavior and Collocation Resampling in Physics-Informed Neural Networks" by Yunona Iwasaki and Ching-Yao Lai. We document the codes used to train PINNs for 1D ice-shelf inverse modeling and to analyze training results over repeated trials.

This repo provides the environments suitable for running our codes on Della-gpu, group workstation, and your local machine

Tools for generating and submitting jobs on Princeton cluster