Menus
Dataset loading
Load curated, remote, or local datasets.
Click the database icon in the main menu to open the dataset modal. There are three ways to get data into Browzarr.
Curated catalog
The modal ships with a browsable catalog of public datasets, split into two groups:
- Zarr stores — e.g. SeasFire Cube, Earth System Data Cube (ESDC), CMIP6 experiments, NASA MUR SST, ARCO-OCEAN
- Icechunk stores — e.g. NOAA HRRR/GFS forecasts, ECMWF AIFS, ERA5 surface reanalysis, Met Office forecasts
Each entry shows a label and a short subtitle. Click one to load its metadata.
Remote store by URL
Paste the URL of a Zarr or Icechunk store:
- Remote Zarr — a URL pointing at the store root (where
.zmetadata/zarr.jsonlives), e.g.https://s3.bgc-jena.mpg.de:9000/misc/seasfire_rechunked.zarr - Remote Icechunk — a URL pointing at the Icechunk repository
The store must allow cross-origin requests from the Browzarr origin (CORS). If a remote store fails to load, CORS is the most common culprit — see Remote data.
Local data
With the offline (npm / Julia) versions you can open data from your own disk:
- Local Zarr — a
.zarrdirectory served locally by the Browzarr app - Local NetCDF — a
.ncfile, decoded in the browser via WebAssembly
See Local data for details and layout requirements.
What you see after loading
Once a dataset is selected, Browzarr parses the store's metadata and shows the available variables. From there, continue with Variable selection.