Going Deeper With Deep Learning: Automatically Tracing Internal Reflection Horizons in Ice Sheets—Methodology and Benchmark Data Set

Abstract Mapping the internal stratigraphy of ice sheets serves a variety of glaciological applications, from the study of past ice flows to current distribution of surface mass balance and melting to contemporary ice dynamics, all of which are crucial for improving future projections of sea level r...

Full description

Saved in:
Bibliographic Details
Main Authors: Hameed Moqadam, Daniel Steinhage, Adalbert Wilhelm, Olaf Eisen
Format: Article
Language:English
Published: Wiley 2025-06-01
Series:Journal of Geophysical Research: Machine Learning and Computation
Subjects:
Online Access:https://doi.org/10.1029/2024JH000493
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849431460861181952
author Hameed Moqadam
Daniel Steinhage
Adalbert Wilhelm
Olaf Eisen
author_facet Hameed Moqadam
Daniel Steinhage
Adalbert Wilhelm
Olaf Eisen
author_sort Hameed Moqadam
collection DOAJ
description Abstract Mapping the internal stratigraphy of ice sheets serves a variety of glaciological applications, from the study of past ice flows to current distribution of surface mass balance and melting to contemporary ice dynamics, all of which are crucial for improving future projections of sea level rise. The method of choice for investigating the internal structure of ice sheets is radio‐echo sounding (RES). Mapping englacial stratigraphy has been carried out so far mostly by time‐consuming manual or semi‐automatic methods. Although reliable, such approaches are not feasible for comprehensive analysis of the wealth of data available. Therefore, (semi‐)automatic mapping of the internal stratigraphy from RES radargrams has been a field of interest for about two decades. Here, our goal is to present a complete pipeline for automatic tracing of internal reflection horizons (IRH) of intermediate to large depths in the ice sheet from radargrams using deep learning. We introduce IRHMapNet, which is a deep learning framework that uses a U‐Net‐based architecture to trace IRHs, based on airborne RES data with preprocessing steps such as noise removal and data augmentation, and postprocessing techniques such as morphological filtering and skeletonization. We use a combination of manually fully‐traced radargrams, results of image processing and thresholding method, and layer slope inference for training U‐Net architectures. We evaluate the successful performance of our approach, also for deep and extended IRH analysis, using a variety of metrics and discuss remaining progress and persistent shortcomings of machine learning approaches. The results of our experiments demonstrate that IRHMapNet successfully achieves the objective with specific architecture and data sets.
format Article
id doaj-art-56ee6d1243e24ca2bfc6bac31af99577
institution Kabale University
issn 2993-5210
language English
publishDate 2025-06-01
publisher Wiley
record_format Article
series Journal of Geophysical Research: Machine Learning and Computation
spelling doaj-art-56ee6d1243e24ca2bfc6bac31af995772025-08-20T03:27:37ZengWileyJournal of Geophysical Research: Machine Learning and Computation2993-52102025-06-0122n/an/a10.1029/2024JH000493Going Deeper With Deep Learning: Automatically Tracing Internal Reflection Horizons in Ice Sheets—Methodology and Benchmark Data SetHameed Moqadam0Daniel Steinhage1Adalbert Wilhelm2Olaf Eisen3Alfred Wegener Institute Helmholtz Centre for Polar and Marine Research Bremerhaven GermanyAlfred Wegener Institute Helmholtz Centre for Polar and Marine Research Bremerhaven GermanyConstructor University Bremen GermanyAlfred Wegener Institute Helmholtz Centre for Polar and Marine Research Bremerhaven GermanyAbstract Mapping the internal stratigraphy of ice sheets serves a variety of glaciological applications, from the study of past ice flows to current distribution of surface mass balance and melting to contemporary ice dynamics, all of which are crucial for improving future projections of sea level rise. The method of choice for investigating the internal structure of ice sheets is radio‐echo sounding (RES). Mapping englacial stratigraphy has been carried out so far mostly by time‐consuming manual or semi‐automatic methods. Although reliable, such approaches are not feasible for comprehensive analysis of the wealth of data available. Therefore, (semi‐)automatic mapping of the internal stratigraphy from RES radargrams has been a field of interest for about two decades. Here, our goal is to present a complete pipeline for automatic tracing of internal reflection horizons (IRH) of intermediate to large depths in the ice sheet from radargrams using deep learning. We introduce IRHMapNet, which is a deep learning framework that uses a U‐Net‐based architecture to trace IRHs, based on airborne RES data with preprocessing steps such as noise removal and data augmentation, and postprocessing techniques such as morphological filtering and skeletonization. We use a combination of manually fully‐traced radargrams, results of image processing and thresholding method, and layer slope inference for training U‐Net architectures. We evaluate the successful performance of our approach, also for deep and extended IRH analysis, using a variety of metrics and discuss remaining progress and persistent shortcomings of machine learning approaches. The results of our experiments demonstrate that IRHMapNet successfully achieves the objective with specific architecture and data sets.https://doi.org/10.1029/2024JH000493radioglaciologydeep learningdeep internal reflection horizon (IRH)radargramAntarctic ice sheetU‐Net architecture
spellingShingle Hameed Moqadam
Daniel Steinhage
Adalbert Wilhelm
Olaf Eisen
Going Deeper With Deep Learning: Automatically Tracing Internal Reflection Horizons in Ice Sheets—Methodology and Benchmark Data Set
Journal of Geophysical Research: Machine Learning and Computation
radioglaciology
deep learning
deep internal reflection horizon (IRH)
radargram
Antarctic ice sheet
U‐Net architecture
title Going Deeper With Deep Learning: Automatically Tracing Internal Reflection Horizons in Ice Sheets—Methodology and Benchmark Data Set
title_full Going Deeper With Deep Learning: Automatically Tracing Internal Reflection Horizons in Ice Sheets—Methodology and Benchmark Data Set
title_fullStr Going Deeper With Deep Learning: Automatically Tracing Internal Reflection Horizons in Ice Sheets—Methodology and Benchmark Data Set
title_full_unstemmed Going Deeper With Deep Learning: Automatically Tracing Internal Reflection Horizons in Ice Sheets—Methodology and Benchmark Data Set
title_short Going Deeper With Deep Learning: Automatically Tracing Internal Reflection Horizons in Ice Sheets—Methodology and Benchmark Data Set
title_sort going deeper with deep learning automatically tracing internal reflection horizons in ice sheets methodology and benchmark data set
topic radioglaciology
deep learning
deep internal reflection horizon (IRH)
radargram
Antarctic ice sheet
U‐Net architecture
url https://doi.org/10.1029/2024JH000493
work_keys_str_mv AT hameedmoqadam goingdeeperwithdeeplearningautomaticallytracinginternalreflectionhorizonsinicesheetsmethodologyandbenchmarkdataset
AT danielsteinhage goingdeeperwithdeeplearningautomaticallytracinginternalreflectionhorizonsinicesheetsmethodologyandbenchmarkdataset
AT adalbertwilhelm goingdeeperwithdeeplearningautomaticallytracinginternalreflectionhorizonsinicesheetsmethodologyandbenchmarkdataset
AT olafeisen goingdeeperwithdeeplearningautomaticallytracinginternalreflectionhorizonsinicesheetsmethodologyandbenchmarkdataset