FirnLearn: A neural network-based approach to firn density modeling in Antarctica

Understanding firn densification is essential for interpreting ice core records, predicting ice sheet mass balance, elevation changes and future sea-level rise. Current models of firn densification on the Antarctic ice sheet (AIS), such as the Herron and Langway (1980) model are either simple semi-e...

Full description

Saved in:
Bibliographic Details
Main Authors: Ayobami Ogunmolasuyi, Colin R. Meyer, Ian McDowell, Megan Thompson-Munson, Ian Baker
Format: Article
Language:English
Published: Cambridge University Press 2025-01-01
Series:Journal of Glaciology
Subjects:
Online Access:https://www.cambridge.org/core/product/identifier/S0022143025000267/type/journal_article
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849716775707475968
author Ayobami Ogunmolasuyi
Colin R. Meyer
Ian McDowell
Megan Thompson-Munson
Ian Baker
author_facet Ayobami Ogunmolasuyi
Colin R. Meyer
Ian McDowell
Megan Thompson-Munson
Ian Baker
author_sort Ayobami Ogunmolasuyi
collection DOAJ
description Understanding firn densification is essential for interpreting ice core records, predicting ice sheet mass balance, elevation changes and future sea-level rise. Current models of firn densification on the Antarctic ice sheet (AIS), such as the Herron and Langway (1980) model are either simple semi-empirical models that rely on sparse climatic data and surface density observations or complex physics-based models that rely on poorly understood physics. In this work, we introduce a deep learning technique to study firn densification on the AIS. Our model, FirnLearn, evaluated on 225 cores, shows an average root-mean-square error of 31 kg m−3 and explained variance of 91%. We use the model to generate surface density and the depths to the $550\,\mathrm{kg\,m}^{-3}$ and $830\,\mathrm{kg\,m}^{-3}$ density horizons across the AIS to assess spatial variability. Comparisons with the Herron and Langway (1980) model at ten locations with different climate conditions demonstrate that FirnLearn more accurately predicts density profiles in the second stage of densification and complete density profiles without direct surface density observations. This work establishes deep learning as a promising tool for understanding firn processes and advancing towards a universally applicable firn model.
format Article
id doaj-art-4f4089d96cdd483e943edef38766cfe3
institution DOAJ
issn 0022-1430
1727-5652
language English
publishDate 2025-01-01
publisher Cambridge University Press
record_format Article
series Journal of Glaciology
spelling doaj-art-4f4089d96cdd483e943edef38766cfe32025-08-20T03:12:53ZengCambridge University PressJournal of Glaciology0022-14301727-56522025-01-017110.1017/jog.2025.26FirnLearn: A neural network-based approach to firn density modeling in AntarcticaAyobami Ogunmolasuyi0https://orcid.org/0000-0002-1202-1541Colin R. Meyer1Ian McDowell2Megan Thompson-Munson3https://orcid.org/0000-0003-4718-193XIan Baker4https://orcid.org/0000-0001-6347-6960Thayer School of Engineering, Dartmouth College, Hanover, NH, USAThayer School of Engineering, Dartmouth College, Hanover, NH, USAGraduate Program of Hydrologic Sciences, University of Nevada, Reno, NV, USAUniversity of Colorado, Boulder, CO, USAThayer School of Engineering, Dartmouth College, Hanover, NH, USAUnderstanding firn densification is essential for interpreting ice core records, predicting ice sheet mass balance, elevation changes and future sea-level rise. Current models of firn densification on the Antarctic ice sheet (AIS), such as the Herron and Langway (1980) model are either simple semi-empirical models that rely on sparse climatic data and surface density observations or complex physics-based models that rely on poorly understood physics. In this work, we introduce a deep learning technique to study firn densification on the AIS. Our model, FirnLearn, evaluated on 225 cores, shows an average root-mean-square error of 31 kg m−3 and explained variance of 91%. We use the model to generate surface density and the depths to the $550\,\mathrm{kg\,m}^{-3}$ and $830\,\mathrm{kg\,m}^{-3}$ density horizons across the AIS to assess spatial variability. Comparisons with the Herron and Langway (1980) model at ten locations with different climate conditions demonstrate that FirnLearn more accurately predicts density profiles in the second stage of densification and complete density profiles without direct surface density observations. This work establishes deep learning as a promising tool for understanding firn processes and advancing towards a universally applicable firn model.https://www.cambridge.org/core/product/identifier/S0022143025000267/type/journal_articledeep learningfirn densificationice sheetsmass balance
spellingShingle Ayobami Ogunmolasuyi
Colin R. Meyer
Ian McDowell
Megan Thompson-Munson
Ian Baker
FirnLearn: A neural network-based approach to firn density modeling in Antarctica
Journal of Glaciology
deep learning
firn densification
ice sheets
mass balance
title FirnLearn: A neural network-based approach to firn density modeling in Antarctica
title_full FirnLearn: A neural network-based approach to firn density modeling in Antarctica
title_fullStr FirnLearn: A neural network-based approach to firn density modeling in Antarctica
title_full_unstemmed FirnLearn: A neural network-based approach to firn density modeling in Antarctica
title_short FirnLearn: A neural network-based approach to firn density modeling in Antarctica
title_sort firnlearn a neural network based approach to firn density modeling in antarctica
topic deep learning
firn densification
ice sheets
mass balance
url https://www.cambridge.org/core/product/identifier/S0022143025000267/type/journal_article
work_keys_str_mv AT ayobamiogunmolasuyi firnlearnaneuralnetworkbasedapproachtofirndensitymodelinginantarctica
AT colinrmeyer firnlearnaneuralnetworkbasedapproachtofirndensitymodelinginantarctica
AT ianmcdowell firnlearnaneuralnetworkbasedapproachtofirndensitymodelinginantarctica
AT meganthompsonmunson firnlearnaneuralnetworkbasedapproachtofirndensitymodelinginantarctica
AT ianbaker firnlearnaneuralnetworkbasedapproachtofirndensitymodelinginantarctica