Institute for Marine and Atmospheric Research (IMAU)

Oceans and Climate

 

Nonlinear Ocean and Climate Dynamics

Introduction

Earth`s climate can be considered as a periodically forced stochastic dynamical system. The periodic forcing is due to the Sun and the Moon (e.g., the tides). The components of the climate system, i.e., atmosphere, ocean, land and ice, and their coupling give rise to internal variability on many temporal and spatial scales. For typical climate phenomena, small-scale variability can be considered as noise. The signal which we see in our observational data and (proxy) records is a result of the evolution of this noisy, periodically forced and (deterministically) complex system.

Climate science is hot and the number of scientists in this field is expected to grow in the near future. There is a strong need for a synthesis of research results as there is much confusion and disagreement on basic mechanisms of climate variability. We are convinced that dynamical systems methodology applied to a hierarchy of models is essential for the development of a scientific theory of the climate system and climate variability. Work within Nonlinear Ocean and Climate Dynamics theme at IMAU is centered around the applications of dynamical systems theory to problems in physical oceanography and climate dynamics.

Contact: Prof. dr. ir. H.A. Dijkstra


Ergodic Theory in the Climate System

Directly related to dynamical systems is the ergodic theory of climate, as a high-dimensional and chaotic system. Indeed, following Lorenz ‘ s assertion “ meteorology is what you get, climate is what you expect “, it has become common practice to describe climate variability as the evolution of statistics. The latter is governed by transfer operators and it has recently become apparent that much could be learned about climate variability and response to forcing from their spectral properties.

Contact: A. Tantet MSc.


Climate Networks

Climate Network derived from Complex Network is an innovative and powerful tool to investigate the patterns and the dynamics of climate variability, because it maps out the topological features that are related to the physics of the dynamical systems. Our research goal is to get better understanding on specific phenomena of climate variability using the techniques from Climate Networks, such as the reduction of the Atlantic Meridional Overturning Circulation, the multidecadal variability associated with the Atlantic Multidecadal Oscillation, and the interannual variations of El NiƱo-Southern Oscillation.

Contact: dr. Q. Feng

Visualization of a climate network. Figure from [Tominski et al., 2011]


Numerical Bifurcation Analysis

To study Earth`s climate as a dynamical system it is vital to obtain models that are capable of easily calculating equilibria and their stability. A powerful approach is to apply numerical continuation techniques to implicitly formulated models, which enables the direct computation of bifurcation diagrams. Essential to this approach is the computation and inversion of Jacobian matrices arising from implicitly formulated physical constraints. Our current goal is to achieve a computational framework in which (parallel) implicit climate models can synchronize and produce coupled Jacobian matrices and suitable coupled preconditioners, giving researchers the opportunity to pursue numerical continuation studies with an Earth system model.

Contact: T.E. Mulder MSc.


The Oceanographic Multipurpose Software Environment (OMUSE)

Multi-scale and multi-physics simulations are an important tool to understand ocean dynamics and its interaction with the different components of the Earth climate system. OMUSE is a framework to facilitate coupling different (mainly oceanographic) simulation codes, building on the AMUSE project developed in the astrophysical community. OMUSE uses the Python programming language to provide homogeneous interfaces to existing numerical codes. The OMUSE framework handles unit conversions, provides consistent and modern object oriented interfaces to the data, manages the state of the underlying simulation codes and provides for transparent distributed computing. OMUSE provides a powerful toolbox to conduct numerical experiments. OMUSE is being developed in collaboration with the eScience Center and Leiden Observatory.

Contact: dr. F.I. Pelupessy