Figure from Sebastian Scher's dissertation

 

Name

Sebastian Scher, PhD candidate
Department of Meteorology, Stockholm University, Sweden

Title

Artificial intelligence in weather and climate prediction: Learning atmospheric dynamics

Abstract

Weather and climate prediction is dominated by high dimensionality, interactions on many different spatial and temporal scales and chaotic dynamics. This makes many problems in the field quite complex ones, and also state-of-the-art numerical models are - despite their immense computational costs - not sufficient for many applications. Therefore, it is appealing to use emerging new technologies such as artificial intelligence to tackle these problems.

We show that it is possible to use deep neural networks to emulate the full dynamics of a strongly simplified general circulation model, providing both good forecasts of the model state several days ahead as well as stable long-term climate timeseries. This method partly also works on more complex and realistic models, but only for forecasting the model's weather several days ahead, not for creating climate runs. It is sufficient to use 50-100 years of data for training the networks. The same neural network method can be combined with singular value decomposition from numerical ensemble weather forecasting in order to generate probabilistic ensemble forecasts with the neural networks.

On a more fundamental level, we show that in a simple dynamical systems setting there seem to be limitations in the ability of feed-forward neural networks to generalize to new regions of the system. This is caused by different parts of the network learning to model different parts of the system. Contradictory, for another simple dynamical system this is shown not to be an issue, raising doubts on the usefulness of results from simple models in the context of more complex ones. Additionally, we show that neural networks are to some extent able to “learn” the influence of slowly changing external forcings on the dynamics of the system, but only given broad enough forcing regimes.
Finally, we present a method to complement operational weather forecasts. Given the initial fields and the error of past weather forecasts, a neural network is used to predict the uncertainty in new forecasts, given only the initial field of the new forecast.

Time and Place

Friday June 12, 2020, at 10.00

Remotely via Zoom.

You are invited to a Zoom webinar.
When: Jun 12, 2020 09:30 AM Paris
Topic: PhD defence Sebastian Scher June 12, 2020

Please click the link below to join the webinar:
https://stockholmuniversity.zoom.us/j/68441165542

Or iPhone one-tap :
    Sweden: +46850520017,,68441165542#  or +46850539728,,68441165542#
Or Telephone:
    Dial(for higher quality, dial a number based on your current location):
        Sweden: +46 8 5052 0017  or +46 850 539 728  or +46 8 4468 2488  or +46 8 5050 0828  or +46 8 5050 0829
        Denmark: +45 32 72 80 11  or +45 89 88 37 88  or +45 32 70 12 06  or +45 32 71 31 57  or +45 32 72 80 10
        Norway: +47 2396 0588  or +47 7349 4877
        Finland: +358 3 4109 2129  or +358 9 4245 1488
        US: +1 312 626 6799  or +1 346 248 7799  or +1 408 638 0968  or +1 408 652 8184  or +1 646 876 9923  or +1 669 900 6833  or +1 253 215 8782  or +1 301 715 8592
    Webinar ID: 684 4116 5542
    International numbers available: https://stockholmuniversity.zoom.us/u/cC2vPeDV0

Or an H.323/SIP room system:
    H.323: 109.105.112.236 or 109.105.112.235
    Meeting ID: 684 4116 5542
    SIP: 68441165542@109.105.112.236 or 68441165542@109.105.112.235