Federal GIS Data Saved and Archived

Federal GIS Data Saved and Archived
Hospitals and Flood Zones from HIFLD Open, Basemap from OpenStreetMap

The following post was written by Frank Donnelly, the Head of GIS and Data Services at the Brown University Library and one of our DRP rescuers!


The Data Rescue Project is pleased to announce that we have archived all of the data layers from the defunct HIFLD Open GIS repository in DataLumos. This collection provides foundational geospatial layers for the nation as a whole. While useful for any type of GIS project, it is particularly important for emergency management, regional planning, and suitability analysis. The geographic features we archived represent both infrastructure and natural features: schools, hospitals, power plants, highways, rivers, flood zones, ports, mines, seismic areas, administrative boundaries, and much more. 

HIFLD Open was a portal that gathered GIS data produced by dozens of federal agencies in one convenient location. Much of the portal’s content is still available from the individual agencies that produced the datasets, and will continue to be updated by these agencies; it will just be more challenging to find and gather. The collection in DataLumos captures a final snapshot of the repository right before it went offline in August of 2025. It will be useful for data users and mappers who need easily accessible and relatively current data for the next couple of years; after that, it will serve primarily as a historic archive.

Work began over the summer of 2025, when two volunteers joined forces to download all the layers in the repository before it was disconnected. One wrote Python scripts that downloaded as much as possible using the portal's API, while the other vetted the results, identified what was missing, and created a comprehensive inventory with metadata for the records. The initiative kicked into high gear in the fall, as twenty volunteers came together and followed a documented process for uploading the data into DataLumos. One volunteer wrote a script that automated metadata entry, which was leveraged by other volunteers to accelerate the upload process. Students working in my lab identified, downloaded, and created extracts of missing datasets, and contributed to the upload process. Ultimately, 415 GIS data layers were archived.

A search for HIFLD Open in DataLumos will yield all of the records, which can be downloaded individually. The titles of the datasets and their descriptions were carried over from the old repository. Users who wish to download the entire set can follow the instructions for connecting to the Brown University Library's Globus endpoint. Globus is a mechanism for transferring big datasets; once you have a Globus account and have installed a small software package, you can search for the collection and transfer everything from a single folder. 

Many thanks to all of the volunteers affiliated with the HIFLD Open initiative at the Data Rescue Project, with special thanks to tripleshrimp, Chiara, and sefk, and to my student workers at GeoData@SciLi.