geozarr

Define GeoZarr-conformant metadata for georeferenced Zarr stores.

66|10|Updated Mar 19, 2026
One-click install
npx skills add https://github.com/isaaccorley/geospatial-skills --skill geozarr
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: geozarr
Source: https://github.com/isaaccorley/geospatial-skills/tree/main/plugins/geozarr/skills/geozarr
Command: npx skills add https://github.com/isaaccorley/geospatial-skills --skill geozarr

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

GeoZarr solves the problem of georeferenced Zarr files that open without a usable CRS/transform, forcing you into repetitive per-tool metadata patching for xarray, rioxarray, and GDAL.

Core Features & Use Cases

  • CF grid_mapping compliance: Encodes CRS and grid mapping in a way that readers can interpret consistently.
  • Correct coordinate semantics: Ensures coordinate arrays use cell centers while GeoTransform uses pixel edges to avoid half-pixel shifts.
  • Zarr v2/v3 compatibility: Supports dimension declarations across Zarr versions so arrays open with the correct axis meaning.
  • Optional interoperability layers: Adds proj:/spatial:/multiscales conventions when needed for broader ecosystem compatibility.
  • In-place patching: Updates only metadata to add CRS/grid mapping without rewriting large data chunks.

Quick Start

Use geozarr when you author a Zarr store from raster inputs and want it to round-trip as a properly georeferenced raster without tool-specific fixes.

Frequently Asked Questions about geozarr

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add CRS metadata to a Zarr store so it opens correctly in GDAL?

Georeferenced Zarr stores often lack a usable CRS or transform, requiring repetitive per-tool metadata patching. GeoZarr solves this by defining CF grid_mapping-compliant metadata with WKT2 and GeoTransform so xarray, rioxarray, and GDAL read them consistently.

Why does my xarray Zarr raster have a half-pixel shift when reading coordinates?

A half-pixel shift in xarray Zarr rasters occurs when coordinate arrays and GeoTransform use mismatched spatial reference points. GeoZarr fixes this by ensuring coordinate arrays use cell centers while GeoTransform uses pixel edges for correct alignment.

Can I patch an existing Zarr store to add grid mapping without rewriting data chunks?

GeoZarr supports both Zarr v2 and v3 by ensuring correct dimension declarations so arrays open with the proper axis meaning. This allows you to author new stores from xarray and rioxarray that round-trip as properly georeferenced rasters.

What is the best way to validate a published Zarr store that lacks CRS on read?

The best way to validate a published Zarr store lacking CRS on read is to check for consistent root Conventions and a valid grid mapping variable. GeoZarr validates these GeoZarr-conformant metadata requirements to restore proper CF grid_mapping behavior.