Deep Learning for He atwaves: From Circulation to Impacts in a Changing Climate
Climate change is causing more and more severe extreme weath er worldwide, such as heat waves. This has major consequences for hea lth, ecosystems and food security. Current models do not always predi ct well, which is why Tamara Happé developed new methods with a focu s on data-driven (AI) models.
Observed temperature tre nds in parts of the Northern Hemisphere, including Western Europe and the Netherlands, are increasing more than the most sophisticated cli mate models had predicted. This is partly due to changes in atmospher ic circulation that the models do not know how to simulate well.
< p>Happé’s research therefore answers three questions: what changes in atmospheric circulation do we see in summer in the Northern Hemis phere, and what are the causes? How can AI methods improve our unders tanding of extreme weather and its associated dynamic changes? And wh at climate patterns do we see during impactful global crop failures?< br>“In addition, I show that data-driven methods, such as
artificial intelligence, can teach us a lot about the climate system
. For example, I show that global crop failures are associated with s
pecific patterns in the ocean and warm and dry conditions over land.
I also show that these specific patterns have become more common in t
he ocean in recent decades.
“Fortunately, such crop failure
s are not common, but can have a lot of impact. The global food indus
try is very dependent on a number of regions in which a lot is produc
ed. If these regions experience crop failures at the same time, it is
noticeable all over the world. With more insight in which climate co
nditions increase the risk of crop failures, and how this changes ove
r time, we can better prepare as a society.”
More information on the thesis.
DESCRIPTION: Deep Learning for Heatwaves: From Circulation to Impacts in a Changing Climate Climate change is causing more and more severe extreme weather worldwide, such as heat waves. This has major consequences for health, ecosystems and food security. Current model s do not always predict well, which is why Tamara Happé developed ne w methods with a focus on data-driven (AI) models.Observed t emperature trends in parts of the Northern Hemisphere, including West ern Europe and the Netherlands, are increasing more than the most sop histicated climate models had predicted. This is partly due to change s in atmospheric circulation that the models do not know how to simul ate well.Happé’s research therefore answers three questions: what changes in atmospheric circulation do we see in summer in the Norther n Hemisphere, and what are the causes? How can AI methods improve our understanding of extreme weather and its associated dynamic changes? And what climate patterns do we see during impactful global crop fai lures?