A Large-Scale Oceanographic Dataset and Prediction Framework for Ocean Currents and Pollution Dispersion

Overview

MOPD-OCPNet consists of a large Marine Oceanic Pollution and Dynamics Dataset (MOPD) that integrates topography, currents, and pollution, and a machine learning model (OCPNet) for accurate prediction.

Required Packages

Ensure the following Python packages are installed to execute the steps:

pip install numpy scipy pandas rasterio xarray matplotlib pillow copernicusmarine

File Structure

env_config/
│
├── Data/
│   ├── Combined/
│   │   ├── combined_environment.nc
│   │   ├── large/
│   ├── GOPAF/
│   ├── gebco_2024_sub_ice_topo_geotiff/
│   ├── Visualizations/
│
├── other_code/
│   ├── buildmap.ipynb
│   ├── gopaf_original.ipynb
│   ├── GeoTIFF_original.ipynb
│   ├── geotifftest2.ipynb
│
├── main.py # Calling the modules
├── Combine.py # [Data Integration Module] Fusion of terrain (GeoTIFF) and current data (NetCDF), harmonization of grids by RBF interpolation, interpolation of current data and storage as NetCDF (combined_environment.nc).
├── GOPAF_Data.py # [Current Data-processing Module] Acquisition, merging and missing value interpolation of Copernicus marine data
├── GOPAF_visual.py  
├── GeoTIFF_Data.py # [Terrain Data-processing Module] Filter, crop, filter and convert GeoTIFF terrain data. 
├── GeoTIFF_visual.py 
├── get_station_id.py # [NOAA Site Acquisition Module] Finds the closest site ID from NOAA to the specified location
├── get_UW_data.py  # [NOAA Data Acquisition Module] Obtain meteorological and oceanographic data from the NOAA Tides and Currents API for specified site ID
│
├── README.md
├── requirements.txt

---

GEBCO 2025 https://www.gebco.net/data_and_products/gridded_bathymetry_data/ ， The terrain data used in this step is based on GEBCO's gridded bathymetric dataset, the GEBCO_2025 Grid.

• DOI: 10.5285/1c44ce99-0a0d-5f4f-e063-7086abc0ea0f

GOPAF https://marine.copernicus.eu/，

• Full Name: Global Ocean Physics Analysis and Forecast
• Abbreviation: GOPAF
• Provider: Copernicus Marine Service (CMEMS)
• DOI: 10.48670/moi-00016

GMB # Global Marine Biodiversity Dataset with Environmental Elements Data Producer: Dou Fangkun
Dataset Index Number: DATA2024000001
DOI: 10.12157/IOC.AS.20240110.001
CSTR: CSTR:33685.11.IOCAS.20240110.002
Data Sharing Mode: Fully Open
Open License: Applicable

NOAA NCEI Microplastics ，用于验证的也是NOAA NCEI的数据

Marine Microplastics Database – National Centers for Environmental Information (NCEI)

Organization: NOAA National Centers for Environmental Information (NCEI)
Product Name: Marine Microplastics
Access Link: NCEI Marine Microplastics
Data Formats: CSV, JSON, GeoJSON
Availability: Public and Open Access
Last Update: Sep 2025 (Product Flier)

MOPD (Marine Oceanic Pollution and Dynamics Dataset)

Pipeline

Step 1: Get Terrain Data and Visualize

1. Dataset Description: Global Ocean & Land Terrain Models
   The terrain data used in this step is based on GEBCO's gridded bathymetric dataset, the GEBCO_2024 Grid. This dataset provides a global terrain model for both ocean and land.

   • Full Name: General Bathymetric Chart of the Oceans 2024 Grid
   • Resolution: 15 arc-second interval grid (approximately 450m at the equator)
   • Elevation Units: Meters
   • Coverage: Global ocean and land terrain
   • Data Format: GeoTIFF files (8 regional files)
   • Data Size: ~15 GB total (8 regional files)
   • Accompanying Data: Includes a Type Identifier (TID) Grid that specifies the source data types used for the GEBCO_2024 Grid
   • Access URL: https://www.gebco.net/data_and_products/gridded_bathymetry_data/
   • DOI: 10.5285/1c44ce99-0a0d-5f4f-e063-7086abc0ea0f

2. Crop Terrain Data

   • File: GeoTIFF_Data.py
   • Function: Get_GeoTIFF_Data
     Extract terrain data based on specified latitude, longitude, and elevation ranges.

3. Visualize Elevation Map

   • File: GeoTIFF_visual.py
   • Function: visualize_elevation_map
     Visualize the cropped terrain data using predefined view angles.

4. Analyze GeoTIFF File (Optional)

   • File: GeoTIFF_Data.py
   • Function: analyze_geotiff
     Analyze terrain data for shape, min/max values, mean, and data type.

5. Convert GeoTIFF to CSV (Optional)

   • File: GeoTIFF_Data.py
   • Function: tif_to_csv
     Convert the terrain data from .tif to .csv for further processing.

---

Step 2: Get Currents Forecast Data (GOPAF)

1. Dataset Description:
   The marine data is sourced from the Global Ocean Physics Analysis and Forecast (GOPAF) datasets provided by CMEMS. We fetch two separate datasets from GOPAF and merge them into a single NetCDF file named combined_gopaf_data.nc.

   • Full Name: Global Ocean Physics Analysis and Forecast
   • Abbreviation: GOPAF
   • Provider: Copernicus Marine Service (CMEMS)
   • Time Range: June 2025 (720 time steps, hourly data)
   • Spatial Coverage: 32°N-33°N, 65.5°W-66.5°W (Boston Harbor area)
   • Depth: 0.494 meters (surface layer)
   • Data Format: NetCDF (CF-1.8 conventions)
   • Variables:
     • Basic Physical Variables:
       • so: Sea Water Salinity (PSU)
       • thetao: Sea Water Potential Temperature (K)
       • uo: Eastward Velocity (m/s)
       • vo: Northward Velocity (m/s)
       • zos: Sea Surface Height Above Geoid (m)
     • Detailed Velocity Variables:
       • utide: Eastward Tidal Velocity (m/s)
       • utotal: Total Eastward Velocity (m/s)
       • vsdx: Stokes Drift X Velocity (m/s)
       • vsdy: Stokes Drift Y Velocity (m/s)
       • vtide: Northward Tidal Velocity (m/s)
       • vtotal: Total Northward Velocity (m/s)
   • Access URL: https://marine.copernicus.eu/
   • DOI: 10.48670/moi-00016
   • References: CMEMS GOPAF Product

2. Fetch and Merge Copernicus Data

   • File: GOPAF_Data.py
   • Function: fetch_and_merge_copernicus_data

3. Complete and Inspect Dataset
   Fill missing values (NaN) using linear and nearest-neighbor interpolation, then verify the structure and attributes for data quality.

Additional Datasets

Global Marine Biodiversity (GMB)

• Full Name: Global Marine Biodiversity Dataset
• Abbreviation: GMB
• Sources: OBIS (Ocean Biodiversity Information System) + GBIF (Global Biodiversity Information Facility)
• Data Scale: 200M+ species occurrence records, 122M sampling points
• Time Span: 1600-2025
• Environmental Variables (13 key parameters):
  • Temperature, Salinity, Sea Currents (ugo, vgo)
  • Chlorophyll, Nitrate, Phosphate, Silicate
  • Dissolved Molecular Oxygen, Net Primary Productivity
  • Dissolved Iron, Carbon Dioxide Partial Pressure, pH Value
  • Phytoplankton Carbon Concentration
• Regional Classification: 8 regions (N_E1, N_E2, N_W1, N_W2, S_E1, S_E2, S_W1, S_W2)
• Access URLs:
  • OBIS
  • GBIF

NOAA Microplastics Data

• Full Name: NOAA National Centers for Environmental Information Microplastics Data
• Provider: NOAA National Centers for Environmental Information (NCEI)
• Data Type: Global ocean microplastics concentration data
• Format: CSV files
• Fields:
  • Latitude/Longitude coordinates
  • Density ranges (1-2, 2-40, 40-200)
  • Density classes (Low, Medium, High)
  • Ocean regions (Atlantic Ocean, etc.)
  • Sampling dates
• Applications: Water quality monitoring, ecosystem protection, remote sensing validation
• Access URL: NOAA NCEI Microplastics

Dataset Integration

• Combined Dataset: combined_environment.nc
• Spatial Resolution: 240×240 grid points
• Time Resolution: 720 hourly time steps
• Variables: 9 ocean physics + 1 terrain variable
• Total Size: >4.2GB integrated dataset

---

Step 3: Combine Terrain and Currents Data

3.1 Build Small Environment

• Interpolate and Merge Data

  • File: Combine.py
  • Function: interpolate_and_merge
    Combine terrain (GeoTIFF) and currents (GOPAF) data into a unified dataset.

• Visualize Combined Data

  • Function: visualize_combined_data
    Visualize the combined dataset by selecting specific time and depth indexes.

3.2 Build Large Environment

• Interpolate Terrain Data

  • Function: interpolate_geotiff
    Increase the resolution of the terrain data for higher accuracy.

• Interpolate and Merge

  • Function: interpolate_and_merge
    Merge the high-resolution terrain data with currents data.

---

(Optional) Fetch Currents Ture Data from NOAA

1. Get Station ID

   • File: get_station_id.py
   • Function: get_nearest_station_id
     Identify the nearest NOAA station for the specified geographic coordinates.

2. Fetch NOAA Data

   • File: get_UW_data.py
   • Function: fetch_noaa_data
     Retrieve the current data using the station ID.

3. Find and Use Current Station

   • File: get_station_id.py
   • Function: find_nearest_current_station
     Determine the closest current station and fetch the corresponding data.

---

OCPNet (Ocean Current and Pollution Prediction Network)

---

Support and References

Data Sources

• E.U. Copernicus Marine Service Information (2024). Global Ocean Physics Analysis and Forecast (GOPAF). doi:https://doi.org/10.48670/moi-00016
• GEBCO Compilation Group (2024). GEBCO 2024 Grid. doi:10.5285/1c44ce99-0a0d-5f4f-e063-7086abc0ea0f
• NOAA National Centers for Environmental Information. Microplastics and oceanographic data. Available at: https://www.ncei.noaa.gov/data/oceans/ncei/microplastics/
• OBIS/GBIF. Global Marine Biodiversity Information System. Available at: https://obis.org/ and https://www.gbif.org/

Dataset URLs

• GEBCO 2024: https://www.gebco.net/data_and_products/gridded_bathymetry_data/
• Copernicus Marine: https://marine.copernicus.eu/
• OBIS: https://obis.org/
• GBIF: https://www.gbif.org/
• NOAA NCEI: https://www.ncei.noaa.gov/

Contact

For questions, support, or collaboration opportunities, please contact the project maintainers.

License

This project is licensed under the MIT License - see the LICENSE file for details.