Meetings and presentations of STERCP
Annual meeting 2019
June 12-14th the annual meeting was held in Bergen, Norway. We presented our work on combining state-of-the-art models, results for synchronized intermediate complex models and neural network supermodeling. Also we enjoyed a couple of invited talks on the wind-driven ocean circulation (Michael Ghil, UCLA) and pseudo-orbit data assimilation (Hailiang Du, Durham University).
SIAM Conference on Applications of Dynamical Systems 2019 (DS19)
At the DS19 meeting we organized the minisymposium Supermodeling an Objective Process by Synchronization of Different Models. Using supermodels formed from climate models and from models of cancer tissue development, this minisymposium addressed issues of training and inter-model coupling, as well as the nonlinearities that make supermodeling superior to ex post facto combination of separate model outputs. Abstracts can be found here .
EGU 2019
At the EGU meeting we organized the session Ensemble Methods for Combining Different Climate (and Weather) Models with several orals and posters about supermodels. Abstracts can be found here .
STERCP meeting December 2018
December 3rd and 4th another STERCP meeting was held in Bergen, Norway. We each gave an update on our work and enjoyed presentations on combining data assimilation and machine learning (Julien Brajard, NERSC) and using contextual model evidence for model selection (Richard Davy, NERSC).
STERCP meeting March 2018
March 22-23 a STERCP meeting was held in Bergen, Norway. We gave an update on our work on combining state-of-the-art models and results for synchronized intermediate complex models, among other things.
Annual meeting 2017
June 26th and June 27th the annual meeting was held in Bergen, Norway. We presented our work on combining state-of-the-art models and results for synchronized intermediate complex models, among other things.
EGU 2017
At the EGU meeting we participated in the session Climate Predictions – from monthly, seasonal to decadal time scales with an oral about efficient training algorithms for supermodels and a poster on SPEEDO supermodels. Abstracts can be found here .
Annual meeting 2016
October 6th and 7th the annual meeting was held in Bergen, Norway. New learning algorithms for supermodels were discussed as well as initial studies with global models.
EGU 2016
At the EGU meeting we participated in the session Climate Predictions – from monthly, seasonal to decadal time scales with an oral about synchronization in the SPEEDO models and posters on Bayesian attractor learning and synchronization of large-scale circulation. Abstracts can be found here.