Atmospheric blocking events can cause severe weather extremes. In a groundbreaking study, scientists used a deep learning model to reconstruct the frequency of these events over the past 1,000 years, offering insights into their future under climate change.
Atmospheric blocking events are prolonged, impactful weather patterns that develop when large high-pressure systems remain stationary, disrupting the jet stream and storm paths for days or even weeks. These events can lead to extreme weather conditions, like record-breaking floods or heat waves, as seen in Europe in 2023.
In a new study, University of Hawai‘i at Mānoa atmospheric scientist Christina Karamperidou used a deep learning model to infer the frequency of blocking events over the past 1,000 years and shed light on how future climate change may impact these significant phenomena.
“This study set out to extract a paleoweather signal from paleoclimate records using a deep learning model that infers atmospheric blocking frequency from surface temperature,” said Karamperidou. “This is a unique study and the first attempt to reconstruct a long record of blocking frequencies based on their relationship with surface temperature, which is complex and unknown. Machine learning methods can be very powerful for such tasks.”
Training the deep learning model
Karamperidou developed a specialized deep learning model, which she trained using historical data and large ensembles of climate model simulations. The model was then capable of inferring the frequency of blocking events from anomalies in seasonal temperature reconstructions over the Last Millennium. These past temperature reconstructions are relatively well-constrained by extensive networks of tree-ring records sensitive to temperature during the growing season.
“This approach demonstrates that deep learning models are powerful tools to overcome the long-standing problem of extracting paleoweather from paleoclimate,” said Karamperidou. “This approach can also be used for the instrumental period of climate history, which began in the 18th century when routine weather measurements were made, since we only have reliable data to identify blocking since the 1940s, or possibly only the satellite era (post-1979).”
Frequency of future blocking events
There isn’t yet a scientific consensus regarding how climate change will change the frequency of blocking events. These strong, persistent mid-latitude high-pressure systems can have significant impacts on Hawai‘i, where flooding has accompanied persistent North Pacific blocks, and also worldwide, for example, in the Pacific Northwest and Europe, where summertime blocking can bring extreme heat waves.
So understanding changes in the frequency of these events, especially as they relate to the other big players for climate, such as El Niño and the long-term patterns of sea surface temperatures in the tropical Pacific, is very important for Hawai‘i. This study allowed Karamperidou to relate blocking frequencies in the mid- and high-latitudes to tropical Pacific climate variability in the long context of the last millennium, which is essential for climate model validation and to narrow down uncertainties in future climate projections of blocking.
Open research and transparency
Karamperidou worked with two UH Mānoa students to create a unique web-interface to explore the deep learning model and the resulting reconstructions. She highlighted that sharing results and methods in this way is important for Open Research best practices and transparency, especially as the application of machine learning and artificial intelligence expands rapidly into many aspects of daily life. The web-interface is hosted on Jetstream-2, an NSF-supported cloud computing system whose regional partners include the University of Hawai‘i Information Technology Services – Cyberinfrastructure and the Hawai‘i Data Science Institute.
In the future, Karamperidou plans to explore a range of features and architectural enhancements of the deep learning model to expand its applications for climate phenomena and variables directly related to high socioeconomic impacts.
Reference: “Extracting paleoweather from paleoclimate through a deep learning reconstruction of Last Millennium atmospheric blocking” by Christina Karamperidou, , 30 September 2024, Communications Earth & Environment.
DOI: 10.1038/s43247-024-01687-y
The study was funded by the U.S. National Science Foundation.
Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.
2 Comments
“This is a unique study and the first attempt to reconstruct a long record of blocking frequencies based on their relationship with surface temperature, which is complex and UNKNOWN.”
In other words, this ‘study’ is a speculative exploration of how well neural nets (NN) work predicting blocking based on proxy tree rings. Tree rings have a low correlation with temperatures because, among other reasons, it is assumed that the correlation is linear. In all probability, the correlation is highest in a range that is generally unknown and probably becomes negative when the optimum is reached. There are other things such as relative humidity, ground water abundance, cloudiness, insect predation, and adjacent plant competition that can impact the correlation. Most of these are unknowable from tree rings alone. Therefore, there are many unstated, let alone studied, assumptions and confounding factors that aren’t addressed. As to caveats addressing the veracity of conclusions, the authors remark, “For most of these regional projections there is no robustness across models, blocking detection algorithms, or climate change scenarios.” I wouldn’t put a lot of stock in the conclusions based on tree-ring proxies. Additionally, one of the intrinsic problems with NNs is that what decisions that are made, and how they are made is difficult to discern. Therefore, predictive systems, such as flying an aircraft, can perform better than a human in real time; however, understanding how they work is generally problematic. That means, developing working hypotheses to guide further research is difficult.
Give us a solid date for a prediction based on that model – then we might start trusting you.