The Instructed Glacier Model (IGM) is an open-source Python package, which permits to simulate 3D glacier evolution accounting for the coupling between ice thermo-dynamics, surface mass balance, and mass conservation. IGM features:
-
Accessibility and modularity: IGM is implemented in the most popular programming language -- Python -- and takes large benefit of existing tools such as OGGM. It is organized module-wise to facilitate coupling, customization and commmunity development. For simplicity, IGM uses a horizontal regular grid for numerical discretization and therefore deals with 2D gridded input and output data.
-
State-of-the-art physics: IGM implements mass conservation, high-order 3D ice flow mechanics, an Enthalpy model for the thermic regime of ice, melt/accumulation surface mass balance model, and other glaciological processes.
-
Computational efficiency: Thanks to the TensorFlow library, mathematical operations are fully-vectorized. This permits tremendous speed-ups on GPU. Physics-informed deep learning is used as an alternative to numerical solvers for modelling ice flow physics in a vectorized way. While GPU are highly-recommended for modelling large domain / high resolution, IGM runs fairly well on CPU for individual glaciers.
-
Automatic differentiation: TensorFlow operations are differentiable. Therefore, automatic differentiation strongly facilitates and speeds-up inverse modelling / data assimilation.
Start with the 10-min video tutorial. Then, all the documentation can be found on the dedicated wiki and the the in-progress and open-to-contributions technical paper.
IGM has a discord channel for quick support, getting latest infos, exchanges with other users. To get in, please send me your discord user name at guillaume.jouvet at unil.ch to be added to the group.
Feel free to drop me an email for any questions, bug reports, or ideas of model extension: guillaume.jouvet at unil.ch