Special Lecture 1: Critical and gradual transitions in pattern-forming systems on Thursday, 21st July 10 am in room D218
Special Lecture 2: Decadal Climate Predictions Using Sequential Learning Algorithms on Monday, 25th July, 3 pm in room D216.
Details
Special Lecture 1: Critical and gradual transitions in pattern-forming systems on Thursday, 21st July 10 am in D218
Critical transitions have attracted a great deal of attention due to their relevance to many natural and social systems. Much research has been devoted to the characterization and identification of imminent critical transitions. In spatially extended systems, the dynamics (close to and away from the critical point) is more complicated due to the expansion, shrinking and coalescence of alternative-state domains. Pattern-forming systems introduce additional complexity due to the patterned nature of one of the stable states. In this talk, I will present several works in which we used the context of drylands vegetation dynamics to study various aspects of this additional complexity: (i) Using a minimal model, we showed that in systems exhibiting a bistability of a patterned state with a uniform state, a multitude of intermediate stable localized states may appear, giving rise to step-like gradual shifts with extended pauses at these states. This result suggests that a combination of abrupt-shift indicators and gradual-shift indicators might be needed to unambiguously identify regime shifts. (ii) The existence of these localized states in models for the dynamics of drylands vegetation and the response of the systems described by these models to local perturbations will be discussed. (iii) We show how a simplified version of a model for drylands vegetation dynamics can explain the emergence and the observed dynamics of the spectacular phenomenon of “fairy circles” in southern Africa.
Special Lecture 2: Decadal Climate Predictions Using Sequential Learning Algorithms on Monday, 25th July, 3 pm in room D216.
Ensembles of climate models are commonly used to improve climate predictions and assess the uncertainties associated with them. Weighting the models according to their performances holds the promise of further improving their predictions. Using an ensemble of climate model simulations from the CMIP5 decadal experiments, we quantified the total uncertainty associated with these predictions and the relative importance of model and internal uncertainties. Sequential learning algorithms (SLAs) were used to reduce the forecast errors and reduce the model uncertainties. The reliability of the SLA predictions was also tested, and the advantages and limitations of the different performance measures are discussed. The spatial distribution of the SLAs performance showed that they are skillful and better than the other forecasting methods over large continuous regions. This finding suggests that, despite the fact that each of the ensemble models is not skillful, the models were able to capture some physical processes that resulted in deviations from the climatology and that the SLAs enabled the extraction of this additional information. If time permits I will also present a method for estimating the uncertainties associated with ensemble predictions and demonstrate the resulting improved reliability.
References:
- Improvement of climate predictions and reduction of their uncertainties using learning algorithms, Atmospheric Chemistry and Physics 15, 8631-8641 (2015).
- Decadal climate predictions using sequential learning algorithms, Journal of Climate 29, 3787-3809 (2016).
- The contribution of internal and model variabilities to the uncertainty in CMIP5 decadal climate predictions, Climate Dynamics 49, 3221 (2017).
- Quantifying the uncertainties in an ensemble of decadal climate predictions. Journal of Geophysical Research: Atmospheres 122, 13,191–13,200 (2017).
- Learning algorithms allow for improved reliability and accuracy of global mean surface temperature projections. Nature Communications 11, 451 (2020).