Preprints
https://doi.org/10.5194/egusphere-2024-2455
https://doi.org/10.5194/egusphere-2024-2455
23 Sep 2024
 | 23 Sep 2024
Status: this preprint is open for discussion.

A gradient-boosted tree framework to model the ice thickness of the World's glaciers (IceBoost v1)

Niccolò Maffezzoli, Eric Rignot, Carlo Barbante, Troels Petersen, and Sebastiano Vascon

Abstract. Knowledge of glacier ice volumes is crucial for constraining future sea level potential, evaluating freshwater resources, and assessing impacts on societies, from regional to global. Motivated by the disparity in existing ice volume estimates, we present IceBoost, a global Machine Learning framework to model individual glacier ice thickness distributions. IceBoost is an ensemble of two gradient-boosted trees trained with 3.7 million globally-available ice thickness measurements and an array of 34 numerical features. The model error is similar to existing models outside polar regions and up to 30–40 % lower at high latitudes. Providing supervision by exposing the model to available glacier thickness measurements reduces the error by up to a factor 2 to 3. A feature ranking analysis reveals that geodetic information are the most informative variables, while ice velocity can improve the model performance by 6 % at high latitudes. A major feature of IceBoost is a capability to generalize outside the training domain, i.e. producing meaningful ice thickness distributions in all regions of the World, including in the ice sheet peripheries.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.
Niccolò Maffezzoli, Eric Rignot, Carlo Barbante, Troels Petersen, and Sebastiano Vascon

Status: open (until 18 Nov 2024)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
Niccolò Maffezzoli, Eric Rignot, Carlo Barbante, Troels Petersen, and Sebastiano Vascon

Data sets

IceBoost - a Gradient Boosted Tree global framework for glacier ice thickness retrieval Niccolò Maffezzoli et al. https://zenodo.org/records/13145836

Model code and software

IceBoost GitHub repository Niccolò Maffezzoli https://github.com/nmaffe/iceboost

Niccolò Maffezzoli, Eric Rignot, Carlo Barbante, Troels Petersen, and Sebastiano Vascon

Viewed

Total article views: 141 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
101 35 5 141 1 1
  • HTML: 101
  • PDF: 35
  • XML: 5
  • Total: 141
  • BibTeX: 1
  • EndNote: 1
Views and downloads (calculated since 23 Sep 2024)
Cumulative views and downloads (calculated since 23 Sep 2024)

Viewed (geographical distribution)

Total article views: 136 (including HTML, PDF, and XML) Thereof 136 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 05 Oct 2024
Download
Short summary
In this work we introduces IceBoost, a machine learning framework to model the ice thickness distribution of all the World's glaciers, with greater accuracy than state-of-the-art methods. The model is trained on 3.7 million measurements globally available and provides skillful estimates across all regions. This advancement will help in better assessing future sea level changes, freshwater resources, with significance for both the scientific community and society at large.