G-RUN ENSEMBLE (pronounced GeRUN) consists in a multi-forcing global reanalysis of monthly runoff rates created by means of machine learning and a global collection of river discharge observations. G-RUN ENSEMBLE allows for an unprecedented view on global terrestrial water dynamics on time scales ranging from months to a full century. Quantification of the uncertainty stemming from the atmospheric forcing data makes G-RUN ENSEMBLE the ideal candidate for reliable and robust water resources assessments.------------------------------------------------------------------------------File description - G-RUN_ENSEMBLE_MMM.nc covers the time period from 1902 to 2019 and provide the median of the G-RUN ENSEMBLE members. If you want to rely on one single estimate this is likely the file you are interested in.- G-RUN_ENSEMBLE_MEMBERS.zip contains ensemble mean reconstructions for 21 different atmospheric forcing datasets. The time range depends on the considered forcing.- Each remaining file called G-RUN_ENSEMBLE_*.zip (where * denotes the acronym of the atmospheric forcing dataset used to force the model), contains 25 runoff reconstructions obtained by training models on different subsets of the available runoff observations. ------------------------------------------------------------------------------References- Ghiggi, G., Humphrey, V., Seneviratne, S. I., & Gudmundsson, L. (2021). G-RUN ENSEMBLE: A multi-forcing observation-based global runoff reanalysis. Water Resources Research, 57(5), e2020WR028787. https://doi.org/10.1029/2020WR028787- Ghiggi, G., Humphrey, V., Seneviratne, S. I., & Gudmundsson, L. (2019). GRUN: an observation-based global gridded runoff dataset from 1902 to 2014. Earth System Science Data, 11(4), 1655���1674. https://doi.org/10.5194/essd-11-1655-2019