Monthly average Global depth-integrated oceanic N₂ fixation rates Dataset (September 1997 - December 2024)

Description

This dataset is a satellite-derived global monthly product of depth-integrated oceanic N₂ fixation rates. The temporal coverage spans from September 1997 to December 2024. Each data file contains N₂ fixation rates (units: µmol N m⁻² d⁻¹) together with corresponding longitude and latitude information, at a spatial resolution of 0.25°. Non-ocean and invalid pixels are assigned as NaN. The data are provided in NetCDF (Network Common Data Form) format. Naming rules: Data files are named following the convention category_YYYYMM_resolution.nc. For example: N2_fixation_199709_025deg.nc, where YYYYMM denotes the year and month, and 025deg indicates a spatial resolution of 0.25°.

Steps to reproduce

This dataset of global monthly depth-integrated oceanic N₂ fixation rates was produced based on multi-source satellite-derived ocean color products and ocean environmental datasets for the period from September 1997 to December 2024. The production process is as follows: Step 1: In situ N₂ fixation rate observations were obtained from the Global Oceanic Diazotroph Database (version 2). Concurrent satellite ocean color data and ocean environmental variables were collected from publicly available global datasets and collocated with the in situ observations in space and time. Step 2: Satellite ocean color variables (e.g., remote sensing reflectance and derived optical properties) and ocean environmental variables (e.g., sea surface temperature, mixed-layer-related parameters, and wind fields) were preprocessed to a common spatial resolution of 0.25° × 0.25° and a monthly temporal scale. Non-ocean and invalid pixels were removed through quality control procedures. Step 3: A machine-learning model was trained using the matched in situ N₂ fixation rates and the corresponding satellite-derived optical and ocean environmental predictors. Model hyperparameters were optimized using cross-validation. Step 4: The trained model was applied to the global satellite ocean color and ocean environmental datasets to generate monthly, depth-integrated N₂ fixation rates for the global ocean. Step 5: The final N₂ fixation fields were post-processed and exported as CF-compliant NetCDF files with longitude and latitude coordinates.

Institutions

• Nanjing University of Information Science and Technology

Categories

Funders

• National Natural Science Foundation of China

  China

  Grant ID: No. 42476173, 42576182, 42176179, 42176181

Licence