Researchers at the University of Alabama used a machine-learning based analysis of extreme floods in the U.S. to discover that these events are associated with four broad atmospheric patterns: tropical moisture exports, tropical cyclones, low-pressure systems, and melting snow.
The researchers used self-organizing maps, a kind of artificial neural network that performs unsupervised clustering, to identify dominant atmospheric circulation patterns linked with extreme floods in the U.S.
The system identified 12 circulation patterns grouped in four major categories.
In addition, the results showed large flood events typically occur in the western and central U.S. because of tropical moisture exports, and in the eastern U.S. because of tropical cyclones.
Weather forecasters, managers of water reservoirs, and emergency management personnel can use the data to better prepare for floods that have a greater likelihood of occurring at the same time as the identified atmospheric patterns.
From University of Alabama
View Full Article
Abstracts Copyright © 2019 SmithBucklin, Washington, DC, USA
No entries found