Ocean acidification, driven by the continuous absorption of atmospheric CO₂, poses severe threats to marine ecosystems and biodiversity. Accurate assessment of seawater pH variations is essential for evaluating biological responses to acidification and predicting the ocean's carbon sequestration capacity.
However, the understanding of global ocean acidification have been limited by sparse seawater pH observations and uneven spatial coverage, particularly below the surface ocean.
To overcome these challenges, the research team from the Institute of Oceanology of the Chinese Academy of Sciences (IOCAS) employed a Stepwise Feed-Forward Neural Network (Stepwise FFNN) algorithm, integrating observational data from the Global Ocean Data Analysis Project (GLODAP) to construct a global monthly 3D gridded pH dataset over the past 30 years.
The study was published in Earth System Science Data on Feb. 24.
"Compared with similar pH products currently limited to surface ocean in the world, our 3D gridded pH dataset extends to the depth of 2000 meters, and has major technical innovations in accuracy and reliability," said Dr. ZHONG Guorong, first author of the study.
By dividing global oceans into biogeochemical provinces based on pH drivers, the optimized selection of environmental variables has enhanced the dataset accuracy. Also, the novel cross-boundary optimal interpolation technology outperforms traditional approaches in the accuracy of reconstructing marine chemical parameters.
The pH dataset has been validated using a rigorous cross-validation method that minimizes model overfitting risks, ensuring reliability. The dataset is openly accessible via the IOCAS Data Center, offering a vital resource for global climate modeling and marine conservation efforts.
"The dataset extending to 2000 meters provides crucial insights into deep ocean acidification processes, supporting advanced studies on ecological responses and climate feedback mechanisms", said Prof. SONG Jinming, the corresponding author.
This achievement marks a methodological leap in marine biogeochemical parameter reconstruction, with potential applications extending to other critical ocean variables. The dataset's high-resolution, full-depth coverage, and robust validation framework position it as an indispensable tool for addressing one of the ocean's most pressing environmental challenges.
Vertical distribution of seawater pH from the 3D gridded dataset. (Image by IOCAS)
(Text by Dr. ZHONG Guorong)
Media Contact:
ZHANG Yiyi
Institute of Oceanology
E-mail: zhangyiyi@qdio.ac.cn
(Editor: ZHANG Yiyi)