DeepMelt-GL v1: a neural network emulator of sub-shelf melt rates for the unrepresented regions of ice-shelf cavities in ocean models
Abstract
The spatial pattern of melting beneath Antarctic ice shelves influences ice flow and retreat, and the resulting freshwater input into the Southern Ocean influences global carbon storage and primary productivity. It is therefore crucial that interactions at the ice-ocean interface are adequately represented in global climate models. However, due to computational limitations, existing climate models are forced to choose between either high resolution or long period simulations, and struggle to resolve melt rates in ice shelf cavities. Here, we show that a simple multilayer perceptron can be used to emulate sub-shelf melt rates in entire small cavities and in the parts of large cavities closest to the grounding zone. We find that the melt rates produced by applying neural network emulators are a good match for melt rates from high resolution simulations, provided that similar conditions were included in the neural network training. We also find that neural networks are particularly sensitive to temperatures and ice-draft slopes outside those used in the training dataset. However, if we train multiple neural networks on the same input data, we demonstrate that the ensemble spread of the neural networks is a good indicator of the reliability of the emulator in any given conditions. The neural network emulator of sub-shelf melt which is presented here, DeepMelt-GL, can be used to improve the representation of ice shelf cavities in both ocean and climate models.
Related
- Drivers of ecological network quality change and multi-scenario simulation and optimization in fragile alpine regions: a case study of Lhasa
- Ocean-Aware Sea-ice Emulator for Future Hybrid Coupled Prediction Systems
- EnsAI: an emulator for atmospheric ammonia ensembles
- Extreme events and statistics in an AI emulator of a global storm-resolving model
- HClimRep-Ocean: A Global Ocean Emulator on an Unstructured Mesh
BibTeX
@article{ockenden2026deepmeltgl,
title = {{DeepMelt-GL v1: a neural network emulator of sub-shelf melt rates for the unrepresented regions of ice-shelf cavities in ocean models}},
author = {Helen Ockenden and Clara Burgard and Pierre Mathiot and Christoph Kittel and Achille Gellens and Cécile Agosta and Nicolas C Jourdain},
journal = {Geoscientific Model Development},
year = {2026},
month = sep,
doi = {10.5194/gmd-19-7979-2026},
url = {https://hal.science/hal-05743340v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - DeepMelt-GL v1: a neural network emulator of sub-shelf melt rates for the unrepresented regions of ice-shelf cavities in ocean models
AU - Helen Ockenden
AU - Clara Burgard
AU - Pierre Mathiot
AU - Christoph Kittel
AU - Achille Gellens
AU - Cécile Agosta
AU - Nicolas C Jourdain
PY - 2026
DA - 2026/09/09
JO - Geoscientific Model Development
DO - 10.5194/gmd-19-7979-2026
UR - https://hal.science/hal-05743340v1
AB - The spatial pattern of melting beneath Antarctic ice shelves influences ice flow and retreat, and the resulting freshwater input into the Southern Ocean influences global carbon storage and primary productivity. It is therefore crucial that interactions at the ice-ocean interface are adequately represented in global climate models. However, due to computational limitations, existing climate models are forced to choose between either high resolution or long period simulations, and struggle to resolve melt rates in ice shelf cavities. Here, we show that a simple multilayer perceptron can be used to emulate sub-shelf melt rates in entire small cavities and in the parts of large cavities closest to the grounding zone. We find that the melt rates produced by applying neural network emulators are a good match for melt rates from high resolution simulations, provided that similar conditions were included in the neural network training. We also find that neural networks are particularly sensitive to temperatures and ice-draft slopes outside those used in the training dataset. However, if we train multiple neural networks on the same input data, we demonstrate that the ensemble spread of the neural networks is a good indicator of the reliability of the emulator in any given conditions. The neural network emulator of sub-shelf melt which is presented here, DeepMelt-GL, can be used to improve the representation of ice shelf cavities in both ocean and climate models.
SN - 1991-9603
SN - 1991-959X
ER -