WorldTensor GeoZarr: 757 harmonized Earth system variables on a 0.25 degree global grid, 1900-2025 (Rodriguez-Pardo and Tavoni, Taylor Geospatial, Source Cooperative)

Provider
Carlos Rodriguez-Pardo and Massimo Tavoni (data); Taylor Geospatial (GeoZarr conversion)
License
CC-BY-4.0 (source)
Format
Zarr (in sampled files)
Stored at
Amazon S3, s3://us-west-2.opendata.source.coop/tge-labs/worldtensor (us-west-2)
Access
Free and public. No account with the data provider is needed.
DOI
10.5281/zenodo.19047618
Category
Environmental Science
Source
https://arxiv.org/abs/2607.03298

About this dataset

WorldTensor is a harmonized collection of 757 Earth system variables regridded to one common 0.25 degree global grid (721 x 1440), mostly annual from 1900 to 2025, assembled by Rodriguez-Pardo and Tavoni as training data for Earth system foundation models. Variables span climate (for example 2-meter air temperature and total precipitation), land use, ocean (such as chlorophyll), cryosphere, hydrology, vegetation, greenhouse gas and pollutant emissions, air quality, energy, natural hazards, human systems, and static soil properties and terrain. This GeoZarr version by Taylor Geospatial stores each variable as its own self-contained Zarr v3 group (worldtensor.zarr/ / , for example climate/t2m_mean or static/soil_properties/bdod_0-5cm_mean) in float32 with Blosc zstd compression, with 2x, 4x and 8x averaged overviews and a ts/ layout that reads a pixel's full history in one request; catalog.json lists every variable with units, years and shape. Values and units are unchanged from the Zenodo release. Open with xarray.open_zarr or rioxarray. Licensed CC BY 4.0, inherited from the Zenodo release (DOI 10.5281/zenodo.19047618); cite Rodriguez-Pardo and Tavoni 2026, arXiv:2607.03298.

How to access this dataset in Python

import boto3
from botocore import UNSIGNED
from botocore.client import Config
s3 = boto3.client('s3', region_name='us-west-2', config=Config(signature_version=UNSIGNED))
s3.list_objects_v2(Bucket='us-west-2.opendata.source.coop', Prefix='tge-labs/worldtensor/', Delimiter='/')

Lists the files with anonymous access. Runs on any machine, with no cloud account.

Data structure

File types in sampled files: Zarr

Files

Example file: tge-labs/worldtensor/worldtensor.zarr/agriculture/fertilizer_nitrogen/2x/fertilizer_nitrogen/c/0/0/0

Analyze this dataset without downloading it

On Scigantic, WorldTensor GeoZarr: 757 harmonized Earth system variables on a 0.25 degree global grid, 1900-2025 (Rodriguez-Pardo and Tavoni, Taylor Geospatial, Source Cooperative) opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.

This page is also available as Markdown and as JSON.

Related Environmental Science datasets