A Global 3D Chlorophyll Concentration Dataset (1997-2024): 4 km Spatial Resolution, Monthly Scale, 0-220 m

Area of Science:

• Oceanography
• Marine Biogeochemistry
• Data Science

Background:

• Oceanic chlorophyll-a (Chl-a) 3D distribution is vital for understanding marine primary productivity, ecosystem dynamics, and the oceanic carbon cycle.
• Existing global 3D Chl-a datasets lack the spatial resolution to capture mesoscale and sub-mesoscale processes.

Purpose of the Study:

• To introduce MEGO-C-M, a high-resolution global 3D Chl-a dataset.
• To provide a dataset with monthly coverage from October 1997 to November 2024 on a 0.0416° grid.
• To enable detailed analysis of oceanic processes at finer scales.

Main Methods:

• Developed the MEGO-C machine-learning framework trained on BGC-Argo profiles and satellite/physical reanalysis predictors.
• Generated gridded Chl-a data across 27 vertical levels using monthly predictor fields.
• Validated the dataset against independent BGC-Argo profiles.

Main Results:

• Achieved strong agreement with independent data (R² = 0.818, RMSE = 0.254 log₁₀(mg·m⁻³), bias = -0.001 log₁₀(mg·m⁻³)).
• The dataset successfully captures vertical Chl-a structures and seasonal variability.
• Ensemble spread is provided per grid cell.

Conclusions:

• MEGO-C-M offers a significant advancement in global 3D Chl-a data resolution.
• The dataset is crucial for resolving mesoscale and sub-mesoscale oceanographic features.
• Available in NetCDF format with a DOI for persistent access.