Earthdata MCP Server

The Earthdata MCP Server bridges the gap between advanced AI assistants and NASA’s vast Earth observation archives. By exposing a lightweight Model Context Protocol interface, it allows Claude or similar agents to programmatically search for and retrieve datasets from the NASA Earthdata portal without leaving the conversational environment. This eliminates the need to manually navigate the web interface, construct complex API calls, or manage authentication tokens—tasks that are typically cumbersome for non‑technical users.

At its core, the server offers two powerful tools: and . The first enables keyword‑based discovery of datasets, supporting optional temporal ranges and geographic bounding boxes to narrow results. The second retrieves individual granules—essentially the raw data files—for a specified dataset, again with filtering on time and space. Both tools return structured lists that can be fed directly into downstream AI reasoning or plotted in notebooks, streamlining workflows from data acquisition to analysis.

Developers find this integration invaluable when building end‑to‑end geospatial pipelines. For example, an AI assistant can ask for “sea level rise datasets from 2000 to 2023,” receive a curated list, then automatically download the relevant granules into a Jupyter session for statistical modeling or visualization. The server’s prompt templates (, ) further simplify common queries, ensuring consistent input formatting and reducing the cognitive load on users.

The Earthdata MCP Server stands out for its seamless orchestration of discovery, retrieval, and analysis within a single AI‑centric workflow. It eliminates the friction of manual data wrangling, accelerates research timelines, and opens high‑resolution Earth observation data to a broader audience—including scientists, educators, and hobbyists—who can leverage AI assistants as knowledgeable guides through the complexities of NASA’s data ecosystem.