Preprints
https://doi.org/10.5194/gmd-2022-57
https://doi.org/10.5194/gmd-2022-57
Submitted as: model evaluation paper
22 Mar 2022
Submitted as: model evaluation paper | 22 Mar 2022
Status: this preprint is currently under review for the journal GMD.

Downscaling Multi-Model Climate Projection Ensembles with Deep Learning (DeepESD): Contribution to CORDEX EUR-44

Jorge Baño-Medina1, Rodrigo Manzanas2, Ezequiel Cimadevilla2, Jesús Fernández1, Jose González-Abad1, Antonio Santiago Cofiño1, and José Manuel Gutiérrez1 Jorge Baño-Medina et al.
  • 1Meteorology Group, Instituto de Física de Cantabria (IFCA,CSIC-UC), Santander, Spain
  • 2Meteorology Group, Dpto. de Matemática Aplicada y Ciencias de la Computación, Universidad de Cantabria, Santander, Spain

Abstract. Deep Learning (DL) has recently emerged as an innovative tool to downscale climate variables from large-scale atmospheric fields under the perfect prognosis (PP) approach. Different Convolutional Neural Networks (CNN) have been applied under present-day conditions with promising results, but little is known about their suitability for extrapolating future climate change conditions. Here, we analyze this problem from a multi-model perspective, developing and evaluating an ensemble of CNN-based downscaled projections (DeepESD) for temperature and precipitation over the European EUR-44i (0.5º) domain, based on eight GCMs from the Coupled Model Intercomparison Project Phase 5 (CMIP5). To our knowledge, this is the first time that CNNs have been used to produce multi-model ensembles of downscaled projections, allowing to quantify inter-model uncertainty in climate change signals. The results are compared with those corresponding to an EUR-44 ensemble of regional climate models (RCMs) showing that DeepESD reduces distributional biases in the historical period. Moreover, the resulting climate change signals are broadly comparable to those obtained with the RCMs, with similar spatial structures. As for the uncertainty of the climate change signal (measured on the basis of inter-model spread), DeepESD yields a smaller uncertainty for precipitation, but a similar uncertainty for temperature.

To facilitate further studies of this downscaling approach we follow FAIR principles and make publicly available the code (a Jupyter notebook) and the DeepESD dataset. In particular, DeepESD is published at the Earth System Grid Federation (ESGF), as the first continental-wide PP dataset contributing to CORDEX (EUR-44).

Jorge Baño-Medina et al.

Status: final response (author comments only)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comments on gmd-2022-57', Anonymous Referee #1, 16 Apr 2022
  • RC2: 'Comment on gmd-2022-57', Anonymous Referee #2, 21 Apr 2022
  • RC3: 'Comment on gmd-2022-57', Anonymous Referee #3, 23 Apr 2022
  • RC4: 'Comment on gmd-2022-57', Anonymous Referee #4, 23 Apr 2022
  • CEC1: 'Comment on gmd-2022-57', Juan Antonio Añel, 25 Apr 2022
  • RC5: 'Comment on gmd-2022-57', Anonymous Referee #5, 06 May 2022

Jorge Baño-Medina et al.

Data sets

DeepESD jorge Baño, Rodrigo Manzanas, Ezequiel Cimadevilla, Jesús Fernández, José González, Antonio Cofiño, José Manuel Gutiérrez https://data.meteo.unican.es/thredds/catalog/esgcet/catalog.html

Model code and software

2022_Bano_GMD.ipynb Jorge Baño, Rodrigo Manzanas, Ezequiel Cimadevilla, Jesús Fernández, José González, Antonio Cofiño, José Manuel Gutiérrez https://github.com/SantanderMetGroup/DeepDownscaling/blob/master/2022_Bano_GMD.ipynb

Jorge Baño-Medina et al.

Viewed

Total article views: 618 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
450 150 18 618 4 5
  • HTML: 450
  • PDF: 150
  • XML: 18
  • Total: 618
  • BibTeX: 4
  • EndNote: 5
Views and downloads (calculated since 22 Mar 2022)
Cumulative views and downloads (calculated since 22 Mar 2022)

Viewed (geographical distribution)

Total article views: 517 (including HTML, PDF, and XML) Thereof 517 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 20 May 2022
Download
Short summary
Artificial intelligence tools, namely deep neural networks, are deployed to produce regional products (i.e., temperature and precipitation) of climate change projections over Europe. The resulting dataset, DeepESD, is make publicly available and analyzed against state-of-the-art methodologies to study the evolution of regional climate, reproducing more accurately the observed climate in the historical period and showing similar trajectories into the future with fewer computational requirements.