The synthetic climate data used in two papers to develop Controlled Abstention Netoworks. The data is approximately 720Mb, saved as a .mat file. The data is from Mamalakis et al. (2021) - with citation given below. Mamalakis, Antonios, Imme Ebert-Uphoff and Elizabeth A. Barnes: Neural Network Attribution Methods for Problems in Geoscience: A Novel Synthetic Benchmark Dataset, submitted to Environmental Data Science, 11/2021, preprint available https://arxiv.org/abs/2103.10005. The code that uses this data can be accessed here: Elizabeth Barnes, & Randal J. Barnes. (2021). eabarnes1010/controlled_abstention_networks: (v1.0.1). Zenodo. https://doi.org/10.5281/zenodo.5750222 The publications associated with this data are posted on arxiv (but will soon be published in JAMES): Barnes, Elizabeth A. and Randal J. Barnes: Controlled abstention neural networks for identifying skillful predictions for regression problems, accepted to JAMES 11/2021. Preprint available at https://arxiv.org/abs/2104.08236 Barnes, Elizabeth A. and Randal J. Barnes: Controlled abstention neural networks for identifying skillful predictions for classification problems, accepted to JAMES 11/2021. Preprint available at https://arxiv.org/abs/2104.08281