04.02.2020
Modern ML methods for climate science should be researched on DKRZ’ supercomputer Mistral, transferred, further developed, and made available to the climate community. One focus is on merging with the modeling of the Earth system, e.g. to improve climate simulations, but also to reconstruct missing climate records.
On February, 26, 2020, Dr. Christopher Kadow is speaker at the MPI-M seminar on "Climate Informatics - New Driver in Climate Science: Artificial Intelligence". The event will take place at 1:30 p.m. in lecture hall H5 of the Geomatikum (Bundesstraße 43, Hamburg). Further information: here
Contact at DKRZ:
Dr. Christopher Kadow
a2Fkb3dAZGtyei5kZQ==
Telephone: +49 40 460094-425
A new AI technology evolved in the new AI/ML group at DKRZ in Cooperation with NVIDIA and the Freie Universität Berlin. By training with climate model data, the CUDA accelerated convolutional neural network is able to reconstruct missing values of e.g. observations.