Original Research ARTICLE
Identifying Dynamic Memory Effects on Vegetation State Using Recurrent Neural Networks
- 1Department of Biogeochemical Integration, Max Planck Institute for Biogeochemistry, Germany
- 2Department of Civil, Geo and Environmental Engineering, Technical University of Munich, Germany
- 3Institute of Data Science, German Aerospace Center, Helmholtz Association of German Research Centers (HZ), Germany
- 4Department of Computer Science, Friedrich Schiller University Jena, Germany
- 5Department of Geography, Friedrich Schiller University Jena, Germany
Vegetation state is largely driven by climate and the complexity of involved processes leads to non-linear interactions over multiple time-scales. Recently, the role of temporally lagged dependencies, so-called memory effects, has been emphasized and studied using data-driven methods, relying on a vast amount of Earth observation and climate data. However, the employed models are often not able to represent the highly non-linear processes and do not represent time explicitly. Thus, data-driven study of vegetation dynamics demands new approaches that are able to model complex sequences. The success of Recurrent Neural Networks (RNNs) in other disciplines dealing with sequential data, such as Natural Language Processing, suggests adoption of this method for Earth system sciences. There are only a few recent studies that used the mentioned models to predict Earth system variables, and none that explore the potential of RNNs as a tool to better understand the temporal scales on which environmental conditions affect vegetation state. Here, we used a Long Short-Term Memory (LSTM) architecture to fit a global model for Normalized Difference Vegetation Index (NDVI), a proxy for vegetation state, by using climate time-series and static variables representing soil properties and land cover as predictor variables. Furthermore, a set of permutation experiments was performed with the objective to identify memory effects and to better understand the scales on which they act under different environmental conditions. This was done by comparing models that have limited access to temporal context, which was achieved through sequence permutation during model training. We performed a cross-validation with spatio-temporal blocking to deal with the auto-correlation present in the data and to increase the generalizability of the findings. With a full temporal model, global NDVI was predicted with R2 of 0.943 and RMSE of 0.056. The temporal model explained 14% more variance than the non-memory model on global level. The strongest differences were found in arid and semiarid regions, where the improvement was up to 25%. Our results show that memory effects matter on global scale, with the strongest effects occurring in sub-tropical and transitional water-driven biomes.
Keywords: memory effects, Lag effects, recurrent neural network (RNN), Long short-term memory (LSTM) network, Normalized difference vegetation index (NDVI)
Received: 18 Apr 2019;
Accepted: 22 Aug 2019.
Copyright: © 2019 Kraft, Martin, Körner, Requena Mesa, Cortés and Reichstein. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Mr. Basil Kraft, Department of Biogeochemical Integration, Max Planck Institute for Biogeochemistry, Jena, Germany, firstname.lastname@example.org