Preprints
https://doi.org/10.5194/essd-2025-273
https://doi.org/10.5194/essd-2025-273
02 Jun 2025
 | 02 Jun 2025
Status: this preprint is currently under review for the journal ESSD.

Reconstructing Global Monthly Ocean Dissolved Oxygen (1960–2023) to Nearly 6000 m Depth Using Bayesian Ensemble Machine Learning

Mingyu Han and Yuntao Zhou

Abstract. Oceanic oxygen levels, crucial for marine ecosystems and biogeochemical cycles, have declined significantly over the past few decades, driven by climate change and posing severe environmental risks. However, historical dissolved oxygen (DO) measurements, especially below 2000 m, remain sparse, limiting comprehensive annual and seasonal analyses. Here we introduce the BEM-DOR framework, a Bayesian-optimized ensemble of six machine-learning models (Random Forest, XGBoost, LightGBM, CatBoost, Extremely Randomized Trees and Histogram-based Gradient Boosting) fused via dynamic weighting, to reconstruct global monthly DO distributions at 1°×1° resolution from the surface to 5902 m depth over 1960–2023. Validation against an independent dataset demonstrates that BEM-DOR outperforms existing products. Our dataset captures depth-dependent deoxygenation, with the most pronounced decline occurring between 150 and 200 m, and reveals dramatically accelerated oxygen loss in the Arctic Ocean and North Atlantic over the past decade. We quantify uncertainties from measurement errors, gridding processes, and model algorithms, providing the first long-term, high-resolution, uncertainty-quantified DO product from ocean surface to nearly 6000 m depth. The extension of DO data into the bathypelagic zone in this work is a significant contribution to deep ocean oxygen dynamics and global biogeochemical cycles.

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.
Share
Mingyu Han and Yuntao Zhou

Status: open (until 09 Jul 2025)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
Mingyu Han and Yuntao Zhou

Data sets

Global Monthly Dissolved Oxygen Reconstruction via Bayesian Ensemble Machine Learning Mingyu Han and Yuntao Zhou https://doi.org/10.5281/zenodo.15361818

Mingyu Han and Yuntao Zhou

Viewed

Total article views: 73 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
64 6 3 73 1 1
  • HTML: 64
  • PDF: 6
  • XML: 3
  • Total: 73
  • BibTeX: 1
  • EndNote: 1
Views and downloads (calculated since 02 Jun 2025)
Cumulative views and downloads (calculated since 02 Jun 2025)

Viewed (geographical distribution)

Total article views: 73 (including HTML, PDF, and XML) Thereof 73 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 07 Jun 2025
Download
Short summary
In recent decades, ocean oxygen levels have fallen steadily, threatening marine life and ecosystem health. To fill gaps in sparse observations, we used an advanced ensemble of six machine learning models to reconstruct monthly oxygen maps from the surface down to 5902 m, covering 1960–2023. Our dataset reveals where and how fast oxygen loss is occurring, especially in deeper waters, and provides a reliable resource for studying climate impacts on oceans and guiding conservation efforts.
Share
Altmetrics