RNN for field reconstruction

Currently, this is the repository for the code of the publication Artificial intelligence achieves easy-to-adapt nonlinear global temperature reconstructions using minimal local data

The code is split into three Jupyter Notebooks:

1. preprocess_netcdf_RNN_public.ipynb

This Notebook deals with processing the raw netcdf input data in order for them to be used for training.

2. LSTM_MPIGE_nonrandom_4git.ipynb

This notebook contains the sampling of the data and the RNN architecture and its execution.

3. postprocess_netcdf_RNN_public.ipynb

This Notebook deals with processing the output of the RNN in order to be compared to other netcdf datasets.

We also added a .csv file with the 140 model architectures evaluated in the testing phase.

---

If you would like to contact us, you can try the following channels

• 📫 fernando.jaume@unige.ch

• Twitter: @fer_jaume

• 📫 martinwegmann@pm.me

• Twitter: @martin_wegmann

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.