MiKlip first phase: MultiCliP

Mechanisms of inter- to multidecadal variability and their implications for climate predictions

MultiCliP will use the comprehensive Earth System Model (ESM) developed at MPI-M in a hierarchy of model set-ups to identify mechanisms and feedbacks determining decadal to multi-decadal-scale variability, and to assess the role of external forcing (e.g., solar, volcanic). MultiCliP will derive implications and recommendations for decadal forecasts.


MultiCliP shall

  • Identify the key processes and feedbacks responsible for decadal to multidecadal variability in long-term MPI-ESM integrations and investigate their specific time scales as well as their regional manifestations.
  • Assess the dependency of these feedbacks and processes on specific model properties, such as resolution and complexity by analyzing existing experiments and dedicated sensitivity experiments.
  • Dissect individual processes, feedbacks, and regions to refine the understanding of the variations and to identify important teleconnections. Dissect coupled feedbacks and teleconnection pathways using idealized atmosphere-ocean forcing and partially coupled experiments.
  • Assess the role of the processes and feedbacks identified above for the predictability of specific climate variables and regions by conducting prototype and idealized prediction experiments.

For these purposes, long-term climate simulations covering the last millennium and unforced control simulations as well as IPCC AR5 20th century simulations and respective control experiments will be analyzed. In addition, a number of sensitivity experiments will be carried out to isolate the effect of resolution in the respective model component and to analyze specific mechanisms.
MultiCliP will thus provide valuable information and recommendations for the configuration of the prototypical prediction system, and will further contribute to the assessment of uncertainty and limits of decadal climate predictability.

MultiCliP - Fig.1
Fig 1: Analysis of variability patterns in ensemble simulations over the last 1200 years: Pacific Decadal Oscillation (PDO) and Atlantic Multidecadal Oscillation (AMO). The red line reflects the AMO reconstruction by Gray et al. (2004). All time series are smoothed by a 31-year running average.

More on MultiCliPNews-Icon


Max-Planck-Institut für Meteorologie
Dr. Johann H. Jungclaus
Dr. Wolfgang Müller
Dr. Jürgen Bader
Dr. Davide Zanchettin