Using machine learning to constrain the atmospheric dynamics contribution to regional climate change

Lead Research Organisation: Imperial College London
Department Name: Physics

Abstract

Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
NE/R011613/1 30/09/2017 30/09/2023
2123640 Studentship NE/R011613/1 30/09/2018 29/06/2022 Carl Thomas
 
Description In my research I have studied atmospheric blocking patterns. These are complex large-scale weather patterns which block the path of the jet stream. They are associated with heat waves in summer and cold snaps in winter. Blocking is poorly understood, and the effect of climate change is not clear. In my research I have developed a new method to study blocking using unsupervised machine learning. I have shown that this method performs better than previous methods used. I have also developed a new data set which has classified these blocking patterns over European summer.

These results show the potential for unsupervised learning in atmospheric science.
Exploitation Route The new blocking index identification and accompanying data set can be used by others. A recent reviewer of the current article has stated that the work "provides a quite "revolutionary" and unique dataset to work with", and this can be used to develop understanding of the long-standing problem of objective blocking identification. This can work towards understanding blocking patterns better, and help solve the long-standing problem in climate science of how climate change is impacting the dynamics of the atmosphere. This long-standing issue is a key source of uncertainty in future projections of extreme events, and solving this issue will lead to better information for climate policy, particularly with regard to adaptation.
Sectors Environment

URL https://wcd.copernicus.org/preprints/wcd-2021-1/
 
Title 100 summers of blocking classification over Europe for UKESM1-0-LL pre-industrial control 
Description A unique classification of regional blocking events in a modern climate model. This has been used to develop and verify a new blocking index that has improved skill at classifying blocking events, particularly in climate models. This can help answer the outstanding question of how blocking events respond to climate change. 
Type Of Material Database/Collection of data 
Year Produced 2021 
Provided To Others? Yes  
Impact This dataset can help answer the outstanding question of how blocking events respond to climate change, and could be used to study the occurrence of blocking events with other methods. 
URL https://doi.org/10.5281/zenodo.4436225
 
Title ERA5 JJA European blocking classification 1979-2019 
Description A unique classification of regional blocking events created from the ERA5 reanalysis for the period 1979-2019, covering European summer. This has been used to develop and verify a new blocking index that has improved skill at classifying blocking events, particularly in climate models. This can help answer the outstanding question of how blocking events respond to climate change. 
Type Of Material Database/Collection of data 
Year Produced 2021 
Provided To Others? Yes  
Impact This dataset can help answer the outstanding question of how blocking events respond to climate change, and could be used to study the occurrence of blocking events with other methods. 
URL https://doi.org/10.5281/zenodo.4436206