We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Using Machine Learning to Analyze Physical Causes of Climate Change: A Case Study of U.S. Midwest Extreme Precipitation.
- Authors
Davenport, Frances V.; Diffenbaugh, Noah S.
- Abstract
While global warming has generally increased the occurrence of extreme precipitation, the physical mechanisms by which climate change alters regional and local precipitation extremes remain uncertain, with debate about the role of changes in the atmospheric circulation. We use a convolutional neural network (CNN) to analyze large‐scale circulation patterns associated with U.S. Midwest extreme precipitation. The CNN correctly identifies 91% of observed precipitation extremes based on daily sea level pressure and 500‐hPa geopotential height anomalies. There is evidence of increasing frequency of extreme precipitation circulation patterns (EPCPs) over the past two decades, although frequency changes are insignificant over the past four decades. Additionally, we find that moisture transport and precipitation intensity during EPCPs have increased. Our approach, which uses deep learning visualization to understand how the CNN predicts EPCPs, advances machine learning as a tool for providing insight into physical causes of changing extremes, potentially reducing uncertainty in future projections. Plain Language Summary: Extreme precipitation and flooding cause widespread impacts on human society. While global warming has increased the occurrence of these damaging events, there is still uncertainty about how climate change will affect precipitation and flooding, making it difficult to adequately prepare for future hazards. We use machine learning to understand why extreme precipitation is becoming more common in the U.S. Midwest by analyzing the atmospheric circulation patterns during extreme precipitation events. Our results show that there is heavier precipitation when extreme precipitation patterns occur, but the patterns themselves have not changed significantly in frequency over the past four decades. Our method could be used to better understand changes in extreme events in the Midwest and in other regions of the world. Key Points: We use a neural network to predict extreme precipitation from daily sea level pressure and 500‐hPa geopotential height fieldsIncreasing Midwest extreme precipitation is linked to increasing moisture transport and precipitation during specific circulation patternsOur method is generalizable to studying how climate change affects the physical causes of various types of extreme events in other regions
- Subjects
MIDWEST (U.S.); MACHINE learning; DEEP learning; ATMOSPHERIC circulation; CONVOLUTIONAL neural networks; GEOPOTENTIAL height
- Publication
Geophysical Research Letters, 2021, Vol 48, Issue 15, p1
- ISSN
0094-8276
- Publication type
Article
- DOI
10.1029/2021GL093787