In the pursuit of understanding the planet's groundwater dynamics, we present GROW (global GROundWater analysis package). This user-friendly, quality-controlled dataset combines groundwater depth and level time series from around the world with associated social-ecological variables. GROW is designed to enable large-sample spatio-temporal groundwater analysis without much further preprocessing.
The dataset contains:
187,317 time series in a daily, monthly, or yearly temporal resolution
from 41 countries – over 90 % of the time series are from either North America, Australia or Europe
most of them are between 10 and 20 years long, from 01/1835 to 04/2024
shallow groundwater tables with a median depth of 8 metres
36 time series or attributes of meteorological, hydrological, geophysical, botanical, and anthropogenic variables (i.e. precipitation, ground elevation, NDVI, irrigated cropland fraction)
32 data flags about well features, as well as time series characteristics (i.e. time series length, trend direction, jumps, autocorrelation, total gap fraction)
Woldmap with location of every time series classified by temporal resolution: On the worldmap, the location of all time series that are included in GROW are shown. The global groundwater data was derived from igrac‘s Global Well and Monitoring Data and Others. The color indicates whether a time series was aggregated to a daily (orange), monthly (green) or yearly (purple) resolution. Additionally, the median groundwater depth per well is given for each temporal resolution. Below, the number of time series available per year (left), the number of time series per length class in years (middle) and the number of time series with no trend, increasing or decreasing trend (right, Mann Kendall trend direction) are displayed (Bäthge et al. in prep. CC BY 4.0).
