Publication in Data Mining and Knowledge Discovery, Steininger et al. (2023)
10/27/2022In collaboration with the Chair of Computer Science X (Data Science) at the University of Würzburg, we have published a paper entitled "ConvMOS: climate model output statistics with deep learning".
Steininger et al. use a convolutional neural network to correct systematic errors in climate models. Several layers of the neural network, which focus on global and regional errors, are used. This methodology has been shown to perform better than established error correction methods with respect to multiple evaluation metrics and to be superior to other machine learning applications.
Click here for the full article:
Steininger, M., Abel, D., Ziegler, K., Krause, A., Paeth, H., and Hotho, A. (2023): “ConvMOS: climate model output statistics with deep learning”. Data Mining and Knowledge Discovery, 37, 136-166. https://doi.org/10.1007/s10618-022-00877-6.
