Earthdata MCP Server: Enhancing Data Discovery and Retrieval for AI Applications

Overview: What is Earthdata MCP Server?

Earthdata MCP Server is a robust implementation of the Model Context Protocol (MCP), specifically tailored to connect AI applications with NASA Earth data resources. This server acts as a bridge, enabling seamless interaction between AI tools such as Claude Desktop, Continue, and Cursor with the vast repository of geospatial datasets available on NASA Earthdata. It provides comprehensive features for dataset discovery and manipulation, making it an essential resource for researchers and developers working in the realm of geospatial analysis.

MCP, akin to the universal adapter USB-C, standardizes communication between AI applications and diverse data sources. This protocol ensures that AI applications can interact with various tools and services without requiring specific custom integrations. The Earthdata MCP Server, therefore, offers a unified interface for accessing NASA Earth data, making it easier for AI developers to incorporate geospatial analysis capabilities into their workflows.

🔧 Core Features & MCP Capabilities

Search Tools

Earthdata MCP Server includes two primary tools designed to facilitate interaction with the data on NASA Earthdata:

1. search_earth_datasets: This tool allows you to search for datasets based on keywords, temporal ranges, and bounding boxes. It returns a list of dataset abstracts that match your criteria.

2. search_earth_datagranules: Similar to search_earth_datasets, this tool searches for data granules using the same parameters but focuses specifically on granule-level information.

These tools enable AI applications to query NASA Earthdata with precision, filtering results based on temporal and spatial constraints.

Integration Depth

Earthdata MCP Server is designed to be deeply integrable with AI applications like Claude Desktop. It supports various client functionalities including resources, prompts, and data access in a standardized manner. This deep integration ensures that AI applications can leverage the vast geospatial dataset catalog of NASA Earthdata effortlessly.

⚙️ MCP Architecture & Protocol Implementation

The implementation of the Model Context Protocol within Earthdata MCP Server follows a well-defined architecture aimed at ensuring robustness and compatibility across different platforms and tools.

Protocol Flow Diagram

graph TD
    A[AI Application] -->|MCP Client| B[MCP Protocol]
    B --> C[MCP Server]
    C --> D[Data Source/Tool]
    style A fill:#e1f5fe
    style C fill:#f3e5f5
    style D fill:#e8f5e8

This diagram illustrates the flow of communication from an AI application (A) to a data source or tool (D) via a standardized MCP protocol. The integration is facilitated through the Earthdata MCP Server, ensuring that any AI application compatible with the MCP can leverage NASA Earthdata efficiently.

Client Compatibility Matrix

MCP Client	Resources	Tools	Prompts	Status
Claude Desktop	✅	✅	✅	Full Support
Continue	✅	✅	✅	Full Support
Cursor	❌	✅	❌	Tools Only

This matrix showcases the compatibility of Earthdata MCP Server with various AI clients. While some clients may have limitations, such as the lack of prompt handling for Cursor, the tool still provides full support for resources and data tools.

🚀 Getting Started with Installation

To start using Earthdata MCP Server, you can either build it from source or pull a pre-built Docker image.

Building from Source

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

This command builds the Docker image for the server, allowing you to run it locally on your machine. This is ideal for developers who wish to make custom modifications.

Pulling Prebuilt Images

make pull-docker

For simplicity, pulling a pre-built image from Docker Hub (datalayer/earthdata-mcp-server) might be more straightforward. Use this command if you prefer not to build the server from source.

docker run -i --rm datalayer/earthdata-mcp-server:latest

💡 Key Use Cases in AI Workflows

Use Case 1: Analyzing Sea Level Rise Trends

Researchers can use Earthdata MCP Server to search for relevant satellite data on sea level rise trends. By specifying temporal and spatial constraints, they can retrieve granular data points, which are then analyzed using specialized geospatial tools within the AI application.

Technical Implementation

from earthdata_mcp_server import search_earth_datagranules

data_granules = search_earth_datagranules(
    short_name="sea_level_rise",
    count=10,
    temporal=("2019-01-01", "2019-12-31"),
    bounding_box=(-75.0, 37.0, -65.0, 45.0)
)

# Use data_granules for further analysis

Use Case 2: Land Cover Classification and Change Detection

Environmental scientists can query the server to find datasets related to land cover change over a specific period. They then use these granule-level data points in their AI models to detect changes and classify different types of land covers.

Technical Implementation

from earthdata_mcp_server import search_earth_datasets

datasets = search_earth_datasets(
    search_keywords="land_cover",
    count=5,
    temporal=("2018-01-01", "2020-12-31"),
    bounding_box=(-90.0, 45.0, -70.0, 60.0)
)

# Use datasets for training and validating land cover classification models

🔌 Integration with MCP Clients

Earthdata MCP Server is fully compatible with popular AI clients like Claude Desktop, Continue, and Cursor. The following integration examples illustrate how to add the server configuration in claude_desktop_config.json.

Configuration Example for Claude Desktop

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

This configuration snippet shows how to add the Earthdata MCP Server as a resource within Claude Desktop. Users can now leverage the powerful geospatial data functionalities directly from their AI application interfaces.

📊 Performance & Compatibility Matrix

Earthdata MCP Server is designed to be highly compatible with various environmental datasets and tools available on NASA Earthdata. The following table provides an overview of its compatibility:

Feature	Support Level
Data Coverage	Comprehensive
Temporal Filtering	High
Spatial Filtering	High
Granule-Level Access	Full

This matrix highlights the robust support for diverse datasets and advanced filtering capabilities.

🛠️ Advanced Configuration & Security

Environmental Variables

Setting environment variables can enhance the security and flexibility of your server setup. For example:

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

Security Best Practices

Ensure that sensitive information such as API keys are stored securely and not hardcoded into your configurations. Additionally, regularly update any dependencies to patch known vulnerabilities.

❓ Frequently Asked Questions (FAQ)

1. Can Earthdata MCP Server be used with other AI clients besides Claude Desktop?

   • Yes, it supports Continue and Cursor, although the compatibility matrix indicates some limitations such as no prompt handling for Cursor.

2. What is the best way to install Earthdata MCP Server?

   • You can either pull a prebuilt Docker image or build from source using make commands.

3. How do I search for specific datasets using temporal and spatial filters?

   • Use the search_earth_datasets tool with parameters like temporal and bounding_box.

4. What is the maximum number of data granules that can be returned in one query?

   • The count parameter allows you to specify how many results are desired, but there is no hard limit specified.

5. How secure is Earthdata MCP Server in handling sensitive data from NASA Earthdata?

   • Environment variables and secure storage practices should be employed to ensure data security when integrating with the server.

👨‍💻 Development & Contribution Guidelines

Contributing to Earthdata MCP Server involves both code contributions and documentation updates. Interested developers can:

1. Clone the repository: git clone https://github.com/datalayer/earthdata-mcp-server
2. Navigate to the working directory: cd earthdata-mcp-server
3. Start contributing by addressing issues or adding new features.

For detailed guidelines, refer to the project's contribution documentation.

🌐 MCP Ecosystem & Resources

Earthdata MCP Server is part of a broader MCP ecosystem that includes various servers and tools designed for AI data access and manipulation. Explore additional resources such as the Model Context Protocol website and other MCP-based servers to enhance your AI application capabilities.

By harnessing the power of Earthdata MCP Server, developers can significantly boost their geospatial analysis workflows, making AI applications more robust and capable.