Training data for neural parameterization of the Lorenz 1980 model in a regime with energetic fast oscillations

The provided data is the solution of the Lorenz 1980 model in the parameter regime specified in Appendix A of the article "The High-Frequency and Rare Events Barriers to Neural Closures of Atmospheric Dynamics" by Mickael D. Chekroun, Honghu Liu, Kaushik Srinivasan, and James C. McWilliams, 2023. The first 1344000 time steps of the data (corresponding to a time window of length 700 days) are used as the training data to obtain the two neural parameterizations called NN1 and NN2 in this article.