Researchers are Using Machine Learning to Better Predict Weather Patterns

Share:

The enormous amount of data from Earth-observing satellites is pushing researchers to use machine learning to mine weather information and improve climate models.  Researchers are using this area of artificial intelligence to discover new climate patterns and understand how climate change is shifting them.

In 2016, researchers first used machine learning to identify tropical cyclones, atmospheric rivers and weather fronts [1].  This same research team is now studying how to use machine learning to find and predict extreme weather patterns and understands shifts in climate.

More: How machine learning could help to improve climate forecasts, Nature, August 23, 2017.

Images correctly identified as tropical cyclones by machine learning. From: Liu et. al, 2016.

Images correctly identified as tropical cyclones by machine learning. From: Liu et. al, 2016.

Reference


Ad:


Liu, Y., Racah, E., Correa, J., Khosrowshahi, A., Lavers, D., Kunkel, K., … & Collins, W. (2016). Application of deep convolutional neural networks for detecting extreme weather in climate datasets. arXiv preprint arXiv:1605.01156.

See Also


Like this article and want more?

Enter your email to receive the weekly GIS Lounge newsletter:

Advertising