A new article titled “A Machine Learning approach for Total Water storage anomaly eXtension back to 1980 (ML-TWiX)” has been published in the journal Nature Scientific Data
The study introduces ML-TWiX, a global dataset of monthly total water storage anomalies (TWSA) reconstructed from 1980 to 2012. The dataset fills an important gap in long-term hydrological records by extending the observational period of TWSA beyond the era of GRACE and GRACE-FO satellite missions, whose combined records span just over two decades.
This continuous reconstruction of TWSA supports a wide range of applications in hydrology, climate science, and water resources assessment, helping researchers study long-term variability and change in terrestrial water storage.
The research was led by GIS (Institute of Geodesy, University of Stuttgart) and includes contributions from an international team of scientists across institutions such as the European Space Agency, ETH Zurich, Jet Propulsion Laboratory, and University of California, Irvine.
The full dataset is publicly available via the University of Stuttgart’s DaRUS repository (https://doi.org/10.18419/DARUS-5233), and the article can be accessed via its DOI: 10.1038/s41597-026-06604-w.
