Preprints
https://doi.org/10.5194/egusphere-2023-1826
https://doi.org/10.5194/egusphere-2023-1826
05 Sep 2023
 | 05 Sep 2023
Status: this preprint has been withdrawn by the authors.

An effective machine learning approach for predicting ecosystem CO2 assimilation across space and time

Piersilvio De Bartolomeis, Alexandru Meterez, Zixin Shu, and Benjamin David Stocker

Abstract. Accurate predictions of environmental controls on ecosystem photosynthesis are essential for understanding the impacts of climate change and extreme events on the carbon cycle and the provisioning of ecosystem services. Using time-series measurements of ecosystem fluxes paired with measurements of meteorological variables from a network of globally distributed sites and remotely sensed vegetation indices, we train a recurrent deep neural network (Long-Short-Term Memory, LSTM), a simple deep neural network (DNN), and a mechanistic, theory-based photosynthesis model with the aim to predict ecosystem gross primary production (GPP). We test these models' ability to spatially and temporally generalise across a wide range of environmental conditions. Both neural network models outperform the theory-based model considering leave-site-out cross-validation (LSOCV). The LSTM model performs best and achieves a mean R2 of 0.78 across sites in the LSOCV and an average R2 of 0.82 across relatively moist temperate and boreal sites. This suggests that recurrent deep neural networks provide a basis for robust data-driven ecosystem photosynthesis modelling in respective biomes. However, limits to global model upscaling are identified using cross-validation by vegetation types and by continents. In particular, our model performance is weakest at relatively arid sites where unknown vegetation exposure to water limitation limits model reliability.

This preprint has been withdrawn.

Piersilvio De Bartolomeis, Alexandru Meterez, Zixin Shu, and Benjamin David Stocker

Interactive discussion

Status: closed

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2023-1826', Anonymous Referee #1, 06 Nov 2023
  • EC1: 'Comment on egusphere-2023-1826 - Peer review halted', Jens-Arne Subke, 14 Dec 2023

Interactive discussion

Status: closed

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2023-1826', Anonymous Referee #1, 06 Nov 2023
  • EC1: 'Comment on egusphere-2023-1826 - Peer review halted', Jens-Arne Subke, 14 Dec 2023
Piersilvio De Bartolomeis, Alexandru Meterez, Zixin Shu, and Benjamin David Stocker
Piersilvio De Bartolomeis, Alexandru Meterez, Zixin Shu, and Benjamin David Stocker

Viewed

Total article views: 542 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
369 155 18 542 12 11
  • HTML: 369
  • PDF: 155
  • XML: 18
  • Total: 542
  • BibTeX: 12
  • EndNote: 11
Views and downloads (calculated since 05 Sep 2023)
Cumulative views and downloads (calculated since 05 Sep 2023)

Viewed (geographical distribution)

Total article views: 524 (including HTML, PDF, and XML) Thereof 524 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 29 Feb 2024
Download

This preprint has been withdrawn.

Short summary
Our research highlights the effectiveness of a recurrent neural network, LSTM, in predicting plant carbon absorption using weather and satellite data. LSTM outperforms other models, even for new locations, suggesting its broad application. Yet, challenges remain in predicting diverse ecosystems globally due to varying plant and climate factors. Our work enhances understanding of Earth's complex ecosystems using advanced models.