Open-source data: A Global-Scale Time Series Dataset for Groundwater Studies within the Earth System

Open-source data: A Global-Scale Time Series Dataset for Groundwater Studies within the Earth System. Paper out this week in Nature from Bäthge et al (2026), introduces GROW (Global-scale integrated GROundWater dataset), a large, standardised, analysis-ready dataset designed to improve understanding of groundwater dynamics within the Earth system.

Groundwater is critical (~99% of accessible freshwater), but its interactions with climate, ecosystems, geology, and human activity are still poorly understood. GROW aims to close this gap with 204,292 groundwater time series from 55 countries.

Temporal resolution:
– 85% yearly
– 9% monthly
– 6% daily

51% of series> 10 years long
36 associated Earth system variables, including:
Climate (precipitation, evapotranspiration)
Hydrology
Geology (rock type)
Biosphere (vegetation, NDVI)
Cryosphere (snow cover)
Human activity (land use, water withdrawal)
34 data-quality flags for filtering (e.g., gaps, trends, outliers)


Paper here: https://www.nature.com/articles/s41597-026-06966-1
Data here: https://zenodo.org/records/15149480
Code:https://github.com/EarthSystemModelling/GROW/blob/main/usage_example.py