Downscaling multi-model climate projection ensembles with deep learning (DeepESD): contribution to CORDEX EUR-44
<p>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 (CNNs) have been applied under present-day conditions with promising results, b...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2022-09-01
|
| Series: | Geoscientific Model Development |
| Online Access: | https://gmd.copernicus.org/articles/15/6747/2022/gmd-15-6747-2022.pdf |
| _version_ | 1811210902469345280 |
|---|---|
| author | J. Baño-Medina R. Manzanas R. Manzanas E. Cimadevilla J. Fernández J. González-Abad A. S. Cofiño J. M. Gutiérrez |
| author_facet | J. Baño-Medina R. Manzanas R. Manzanas E. Cimadevilla J. Fernández J. González-Abad A. S. Cofiño J. M. Gutiérrez |
| author_sort | J. Baño-Medina |
| collection | DOAJ |
| description | <p>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 (CNNs) 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 (hereafter DeepESD) for temperature and precipitation over the European EUR-44i (0.5<span class="inline-formula"><sup>∘</sup></span>) domain, based on eight global circulation models (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 downscaled multi-model ensembles based on the perfect-prognosis approach, allowing us 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 preserves the uncertainty for temperature and results in a reduced uncertainty for precipitation.</p>
<p>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).</p> |
| first_indexed | 2024-04-12T05:03:48Z |
| format | Article |
| id | doaj.art-895eb5e96c694f88a731879dd499a1f5 |
| institution | Directory Open Access Journal |
| issn | 1991-959X 1991-9603 |
| language | English |
| last_indexed | 2024-04-12T05:03:48Z |
| publishDate | 2022-09-01 |
| publisher | Copernicus Publications |
| record_format | Article |
| series | Geoscientific Model Development |
| spelling | doaj.art-895eb5e96c694f88a731879dd499a1f52022-12-22T03:46:55ZengCopernicus PublicationsGeoscientific Model Development1991-959X1991-96032022-09-01156747675810.5194/gmd-15-6747-2022Downscaling multi-model climate projection ensembles with deep learning (DeepESD): contribution to CORDEX EUR-44J. Baño-Medina0R. Manzanas1R. Manzanas2E. Cimadevilla3J. Fernández4J. González-Abad5A. S. Cofiño6J. M. Gutiérrez7Instituto de Física de Cantabria (IFCA), CSIC–Universidad de Cantabria, Santander, SpainDepartamento de Matemática Aplicada y Ciencias de la Computación (MACC), Universidad de Cantabria, Santander, SpainGrupo de Meteorología y Computación, Universidad de Cantabria, Unidad Asociada al CSIC, Santander, SpainInstituto de Física de Cantabria (IFCA), CSIC–Universidad de Cantabria, Santander, SpainInstituto de Física de Cantabria (IFCA), CSIC–Universidad de Cantabria, Santander, SpainInstituto de Física de Cantabria (IFCA), CSIC–Universidad de Cantabria, Santander, SpainInstituto de Física de Cantabria (IFCA), CSIC–Universidad de Cantabria, Santander, SpainInstituto de Física de Cantabria (IFCA), CSIC–Universidad de Cantabria, Santander, Spain<p>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 (CNNs) 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 (hereafter DeepESD) for temperature and precipitation over the European EUR-44i (0.5<span class="inline-formula"><sup>∘</sup></span>) domain, based on eight global circulation models (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 downscaled multi-model ensembles based on the perfect-prognosis approach, allowing us 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 preserves the uncertainty for temperature and results in a reduced uncertainty for precipitation.</p> <p>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).</p>https://gmd.copernicus.org/articles/15/6747/2022/gmd-15-6747-2022.pdf |
| spellingShingle | J. Baño-Medina R. Manzanas R. Manzanas E. Cimadevilla J. Fernández J. González-Abad A. S. Cofiño J. M. Gutiérrez Downscaling multi-model climate projection ensembles with deep learning (DeepESD): contribution to CORDEX EUR-44 Geoscientific Model Development |
| title | Downscaling multi-model climate projection ensembles with deep learning (DeepESD): contribution to CORDEX EUR-44 |
| title_full | Downscaling multi-model climate projection ensembles with deep learning (DeepESD): contribution to CORDEX EUR-44 |
| title_fullStr | Downscaling multi-model climate projection ensembles with deep learning (DeepESD): contribution to CORDEX EUR-44 |
| title_full_unstemmed | Downscaling multi-model climate projection ensembles with deep learning (DeepESD): contribution to CORDEX EUR-44 |
| title_short | Downscaling multi-model climate projection ensembles with deep learning (DeepESD): contribution to CORDEX EUR-44 |
| title_sort | downscaling multi model climate projection ensembles with deep learning deepesd contribution to cordex eur 44 |
| url | https://gmd.copernicus.org/articles/15/6747/2022/gmd-15-6747-2022.pdf |
| work_keys_str_mv | AT jbanomedina downscalingmultimodelclimateprojectionensembleswithdeeplearningdeepesdcontributiontocordexeur44 AT rmanzanas downscalingmultimodelclimateprojectionensembleswithdeeplearningdeepesdcontributiontocordexeur44 AT rmanzanas downscalingmultimodelclimateprojectionensembleswithdeeplearningdeepesdcontributiontocordexeur44 AT ecimadevilla downscalingmultimodelclimateprojectionensembleswithdeeplearningdeepesdcontributiontocordexeur44 AT jfernandez downscalingmultimodelclimateprojectionensembleswithdeeplearningdeepesdcontributiontocordexeur44 AT jgonzalezabad downscalingmultimodelclimateprojectionensembleswithdeeplearningdeepesdcontributiontocordexeur44 AT ascofino downscalingmultimodelclimateprojectionensembleswithdeeplearningdeepesdcontributiontocordexeur44 AT jmgutierrez downscalingmultimodelclimateprojectionensembleswithdeeplearningdeepesdcontributiontocordexeur44 |