bask0/dl4es_ch18

Contains experiments conducted for the DL4ES book, chapter 18. We use an LSTM (and contrast to a feed-forward model) to emulate a land-surface model to demonstrate the capability of an LSTM to capture ecological memory effects on the global scale for various conditions.

Emulating Ecological Memory with Recurrent Neural Networks

The DL4ES book - Chapter 18

Authors
B. Kraft [bkraft@bgc-jena.mpg.de]
S. Besnard [sbesnard@bgc-jena.mpg.de]
S. Koirala [skoirala@bgc-jena.mpg.de]

Info

The repository contains the code for the experiments conducted for the book chapter:

Emulating Ecological Memory with Recurrent Neural Networks in the the DL4ES book

We cannot share the datasets used in the experiment publicly, but we share the code for transparency. Contact us if you need access to the data or simulations. We are happy to help / collaborate!

If you need assistance with the code, contact: bkraft@bgc-jena.mpg.de

Links

• DL4ES book repo
• The Book

Citation

This chapter:

@incollection{Kraft2021emulating,
    title = {Emulating Ecological Memory with Recurrent Neural Networks},
    booktitle = {Deep Learning for the Earth Sciences: A Comprehensive Approach to Remote Sensing, Climate Science and Geosciences},
    author = {Kraft, Basil and Besnard, Simon and Koirala, Sujan},
    editor = {Camps-Valls, Gustau and Tuia, Devis and Zhu, Xiao Xiang and Reichstein, Markus},
    edition = {1st edition},
    publisher = {{Wiley \& Sons}},
    isbn = {978-1-119-64614-3},
    Year = {2021}
}

The book:

@book{CampsValls21wiley,
    title = {Deep learning for the Earth Sciences: A comprehensive approach to remote sensing, climate science and geosciences},
    author = {Camps-Valls, G. and Tuia, D. and Zhu, X.X. and Reichstein, M.},
    edition = {1st edition},
    publisher = {Wiley \& Sons},
    isbn = {978-1-119-64614-3},
    year = {2021}
}