🪐 ✨ Earthdata MCP Server

Earthdata MCP Server is a Model Context Protocol (MCP) server implementation that provides tools to interact with NASA Earth Data. It enables efficient dataset discovery and retrieval for Geospatial analysis.

🚀 NEW: This server now includes all Jupyter MCP Server tools through composition, providing a unified interface for both Earth data discovery and Jupyter notebook manipulation.

The following demo uses this MCP server to search for datasets and data granules on NASA Earthdata, the jupyter-earth-mcp-server to download the data in Jupyter and the jupyter-mcp-server to run further analysis.

Analyzing Sea Level Rise with AI-Powered Geospatial Tools and Jupyter - Watch Video

Use with Claude Desktop

To use this with Claude Desktop, add the following to your claude_desktop_config.json.

{
  "mcpServers": {
    "earthdata": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "datalayer/earthdata-mcp-server:latest"
      ]
    }
  }
}

If you are using Linux, start Claude with the following command.

make claude-linux

Tools

The server offers 15 tools total: 3 Earthdata-specific tools plus 12 Jupyter notebook manipulation tools (prefixed with jupyter_).

Earthdata Tools

search_earth_datasets

• Search for datasets on NASA Earthdata.
• Input:
  • search_keywords (str): Keywords to search for in the dataset titles.
  • count (int): Number of datasets to return.
  • temporal (tuple): (Optional) Temporal range in the format (date_from, date_to).
  • bounding_box (tuple): (Optional) Bounding box in the format (lower_left_lon, lower_left_lat, upper_right_lon, upper_right_lat).
• Returns: List of dataset abstracts.

search_earth_datagranules

• Search for data granules on NASA Earthdata.
• Input:
  • short_name (str): Short name of the dataset.
  • count (int): Number of data granules to return.
  • temporal (tuple): (Optional) Temporal range in the format (date_from, date_to).
  • bounding_box (tuple): (Optional) Bounding box in the format (lower_left_lon, lower_left_lat, upper_right_lon, upper_right_lat).
• Returns: List of data granules.

download_earth_data_granules 🆕

• Download Earth data granules from NASA Earth Data and integrate with Jupyter notebooks.
• This tool combines earthdata search capabilities with jupyter notebook manipulation to create a seamless download workflow.
• Input:
  • folder_name (str): Local folder name to save the data.
  • short_name (str): Short name of the Earth dataset to download.
  • count (int): Number of data granules to download.
  • temporal (tuple): (Optional) Temporal range in the format (date_from, date_to).
  • bounding_box (tuple): (Optional) Bounding box in the format (lower_left_lon, lower_left_lat, upper_right_lon, upper_right_lat).
• Returns: Success message with download code preparation details.
• Integration: Uses composed jupyter tools to add download code to notebooks for interactive execution.

Jupyter Tools (Composed)

The following Jupyter notebook manipulation tools are available with the jupyter_ prefix:

• jupyter_append_markdown_cell: Add markdown cells to notebooks
• jupyter_insert_markdown_cell: Insert markdown cells at specific positions
• jupyter_overwrite_cell_source: Modify existing cell content
• jupyter_append_execute_code_cell: Add and execute code cells
• jupyter_insert_execute_code_cell: Insert and execute code cells at specific positions
• jupyter_execute_cell_with_progress: Execute cells with progress monitoring
• jupyter_execute_cell_simple_timeout: Execute cells with timeout
• jupyter_execute_cell_streaming: Execute cells with streaming output
• jupyter_read_all_cells: Read all notebook cells
• jupyter_read_cell: Read specific notebook cells
• jupyter_get_notebook_info: Get notebook metadata
• jupyter_delete_cell: Delete notebook cells

For detailed documentation of the Jupyter tools, see the Jupyter MCP Server documentation.

Architecture: Server Composition

This server uses a composition pattern to combine tools from multiple MCP servers into a single unified interface. The implementation:

1. Imports the Jupyter MCP Server at runtime
2. Merges tool definitions from the Jupyter server into the Earthdata server
3. Prefixes Jupyter tools with jupyter_ to avoid naming conflicts
4. Preserves all functionality from both servers

This approach provides several benefits:

• ✅ Unified Interface: Single MCP server for both Earth data and Jupyter operations
• ✅ No Duplication: Reuses existing Jupyter MCP Server code without copying
• ✅ Namespace Safety: Prefixed tools prevent naming conflicts
• ✅ Graceful Degradation: Falls back to Earthdata-only if Jupyter server unavailable
• ✅ Maintainability: Changes to Jupyter MCP Server are automatically included

Implementation Details

The composition is implemented in the _compose_jupyter_tools() function, which:

# Simplified version of the composition logic
def _compose_jupyter_tools():
    jupyter_mcp_module = importlib.import_module("jupyter_mcp_server.server")
    jupyter_mcp_instance = jupyter_mcp_module.mcp
    
    # Add jupyter tools with prefixed names
    for tool_name, tool in jupyter_mcp_instance._tool_manager._tools.items():
        prefixed_name = f"jupyter_{tool_name}"
        mcp._tool_manager._tools[prefixed_name] = tool

This pattern can be extended to compose additional MCP servers as needed.

Prompts

1. download_analyze_global_sea_level 🆕

   • Generate a comprehensive workflow for downloading and analyzing Global Mean Sea Level Trend dataset.
   • Uses both earthdata download tools and jupyter analysis capabilities.
   • Returns: Detailed prompt for complete sea level analysis workflow.

2. sealevel_rise_dataset

   • Search for datasets related to sea level rise worldwide.
   • Input:
     • start_year (int): Start year to consider.
     • end_year (int): End year to consider.
   • Returns: Prompt correctly formatted.

3. ask_datasets_format

   • To ask about the format of the datasets.
   • Returns: Prompt correctly formatted.

Building

# or run `docker build -t datalayer/earthdata-mcp-server .`
make build-docker

If you prefer, you can pull the prebuilt images.

make pull-docker