Mitigating Non-Stationarity in Machine Learning-Based Downscaling of Climate Projections
Published in Journal of Advances in Modeling Earth Systems, 2026
Recommended citation: Wang, Y., Visioni, D., Kravitz, B., MacMartin, D. G., and Balwada, D. (2026). Mitigating non-stationarity in machine learning-based downscaling of climate projections. Journal of Advances in Modeling Earth Systems, 18, e2026MS005891. https://doi.org/10.1029/2026MS005891 https://doi.org/10.1029/2026MS005891
‘A study of how non-stationarity degrades machine learning downscaling of temperature and precipitation across emission pathways and two solar radiation modification scenarios in MPI-ESM, showing that a more stationary prediction target improves generalization’
