Artificial intelligence achieves easy-to-adapt nonlinear global temperature reconstructions using minimal local data
Abstract Understanding monthly-to-annual climate variability is essential for adapting to future climate extremes. Key ways to do this are through analysing climate field reconstructions and reanalyses. However, producing such reconstructions can be limited by high production costs, unrealistic line...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-06-01
|
| Series: | Communications Earth & Environment |
| Online Access: | https://doi.org/10.1038/s43247-023-00872-9 |
| _version_ | 1797801229107068928 |
|---|---|
| author | Martin Wegmann Fernando Jaume-Santero |
| author_facet | Martin Wegmann Fernando Jaume-Santero |
| author_sort | Martin Wegmann |
| collection | DOAJ |
| description | Abstract Understanding monthly-to-annual climate variability is essential for adapting to future climate extremes. Key ways to do this are through analysing climate field reconstructions and reanalyses. However, producing such reconstructions can be limited by high production costs, unrealistic linearity assumptions, or uneven distribution of local climate records. Here, we present a machine learning-based non-linear climate variability reconstruction method using a Recurrent Neural Network that is able to learn from existing model outputs and reanalysis data. As a proof-of-concept, we reconstructed more than 400 years of global, monthly temperature anomalies based on sparse, realistically distributed pseudo-station data and show the impact of different training data sets. Our reconstructions show realistic temperature patterns and magnitude reproduction costing about 1 hour on a middle-class laptop. We highlight the method’s capability in terms of mean statistics compared to more established methods and find that it is also suited to reconstruct specific climate events. This approach can easily be adapted for a wide range of regions, periods and variables. |
| first_indexed | 2024-03-13T04:47:16Z |
| format | Article |
| id | doaj.art-3632d8022f74404b8cfa2107013a29bf |
| institution | Directory Open Access Journal |
| issn | 2662-4435 |
| language | English |
| last_indexed | 2024-03-13T04:47:16Z |
| publishDate | 2023-06-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Communications Earth & Environment |
| spelling | doaj.art-3632d8022f74404b8cfa2107013a29bf2023-06-18T11:25:46ZengNature PortfolioCommunications Earth & Environment2662-44352023-06-014111210.1038/s43247-023-00872-9Artificial intelligence achieves easy-to-adapt nonlinear global temperature reconstructions using minimal local dataMartin Wegmann0Fernando Jaume-Santero1Institute of Geography, University of BernDepartment of Radiology and Medical Informatics, University of GenevaAbstract Understanding monthly-to-annual climate variability is essential for adapting to future climate extremes. Key ways to do this are through analysing climate field reconstructions and reanalyses. However, producing such reconstructions can be limited by high production costs, unrealistic linearity assumptions, or uneven distribution of local climate records. Here, we present a machine learning-based non-linear climate variability reconstruction method using a Recurrent Neural Network that is able to learn from existing model outputs and reanalysis data. As a proof-of-concept, we reconstructed more than 400 years of global, monthly temperature anomalies based on sparse, realistically distributed pseudo-station data and show the impact of different training data sets. Our reconstructions show realistic temperature patterns and magnitude reproduction costing about 1 hour on a middle-class laptop. We highlight the method’s capability in terms of mean statistics compared to more established methods and find that it is also suited to reconstruct specific climate events. This approach can easily be adapted for a wide range of regions, periods and variables.https://doi.org/10.1038/s43247-023-00872-9 |
| spellingShingle | Martin Wegmann Fernando Jaume-Santero Artificial intelligence achieves easy-to-adapt nonlinear global temperature reconstructions using minimal local data Communications Earth & Environment |
| title | Artificial intelligence achieves easy-to-adapt nonlinear global temperature reconstructions using minimal local data |
| title_full | Artificial intelligence achieves easy-to-adapt nonlinear global temperature reconstructions using minimal local data |
| title_fullStr | Artificial intelligence achieves easy-to-adapt nonlinear global temperature reconstructions using minimal local data |
| title_full_unstemmed | Artificial intelligence achieves easy-to-adapt nonlinear global temperature reconstructions using minimal local data |
| title_short | Artificial intelligence achieves easy-to-adapt nonlinear global temperature reconstructions using minimal local data |
| title_sort | artificial intelligence achieves easy to adapt nonlinear global temperature reconstructions using minimal local data |
| url | https://doi.org/10.1038/s43247-023-00872-9 |
| work_keys_str_mv | AT martinwegmann artificialintelligenceachieveseasytoadaptnonlinearglobaltemperaturereconstructionsusingminimallocaldata AT fernandojaumesantero artificialintelligenceachieveseasytoadaptnonlinearglobaltemperaturereconstructionsusingminimallocaldata |