11-04, 11:30–12:00 (America/New_York), Regency Ballroom B
Zarr is gaining traction in geospatial workflows—but is it replacing COG, complementing it, or something else entirely? We’ll unpack the formats’ shared foundations, explore their tradeoffs, and offer a path toward better community guidance, tooling, and support.
Cloud-Optimized GeoTIFF (COG) and Zarr have each earned their place in modern geospatial workflows. While often framed in opposition—raster vs. analysis, imagery vs. data cube—they are in fact deeply complementary. In this talk, we’ll unpack how they address similar challenges from different angles, and why they should be considered parts of a shared toolkit rather than competing paradigms.
We’ll highlight where Zarr and COG overlap, where they differ, and how decisions around chunking, compression, tiling strategies, and metadata design affect both formats. We'll discuss implementation pitfalls, emerging best practices, and the still-unanswered questions that data producers and tool builders face.
More than a comparison, this talk is a call to action: the community lacks clear guidance and consistent support for practitioners working to produce data in either format. We’ll highlight concrete gaps in the tooling landscape, share ideas from our own work on how to improve decision-making and best practices, and invite others to collaborate on building a healthier, more cooperative open geospatial data ecosystem.
Jarrett Keifer is a Senior Geospatial Software Engineer at Element 84, a commercial geospatial consultancy that uses open-source to build effective customer solutions. His interests include education and outreach, geospatial data formats, and high-performance systems/network programming. He enjoys designing systems to operate at scale, particularly to support remote sensing data processing and earth science applications, and has over ten years of experience contributing to open source projects.
I have worked in the scientific Python ecosystem as an environmental researcher, an open source contributor, and a web developer. I am passionate about finding creative ways to enhance understanding of the physical world. My past experience includes maintaining Dask (open source distributed computing tool) and HoloViz (open source high-level visualization tool). In my current role at Element 84 (formerly Azavea), I work on the maintain django/react web applications and push forward tooling and open source best practices for scientists.