Small print

This document is currently maintained by Philip Brohan (philip.brohan @ metoffice.gov.uk). All criticism should be directed to him - put please don’t send email, raise an issue instead.

All blame should go to the maintainer; credit is more widely distributed:

• This document was written by Philip Brohan (Met Office). He was funded by the Joint BEIS and Defra Integrated Climate Programme, DECC/Defra (GA01101), and by the Climate Science for Service Partnership China.
• This work follows on from a previous project on weather forecasts with machine learning, and a previous project on `Improving reanalysis with offline assimilation.
• The TensorFlow library is used throughout.
• The 20th Century Reanalysis dataset (version 2c) provides the training data used.
• Vikram Tiwari’s tf.keras Autoencoder page was an invaluable introduction to both autoencoders and Tensorflow.
• Ashley Jager’s weather icons are used in the logo image.
• Background information was provided by research papers, notably:
  • Isola et al. Image-to-Image Translation with Conditional Adversarial Networks
  • Goodfellow et al. Generative Adversarial Nets
  • Mirza & Osindero Conditional Generative Adversarial Nets
  • Radford et al. Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
  • Springenberg et al. Striving for Simplicity: The All Convolutional Net

Note that appearance on this list does not mean that the person or organisation named endorses this work, agrees with any of it, or even knows of its existence.

This document and the data associated with it are crown copyright (2019) and licensed under the terms of the Open Government Licence. All code included is licensed under the terms of the GNU Lesser General Public License.