Work Package 2: Global dynamics of climate variability and change

The objective of this work package (WP) is to understand the dynamics of global climate variability and change, including regional climate predictability and underpinning effective advice and mitigation.

This work package is improving our understanding of the mechanisms and predictability of patterns involved in global climate variation, focussing on how these affect China in particular. This involves developing tools to evaluate how well models simulate observed variability, alongside assessing how well the patterns are predicted on monthly to decadal timescales, which underpins new climate prediction services. The partnership has already developed a prototype climate service for seasonal forecasting of rainfall in the Yangtze River basin.


Dynamical variability has a strong control of regional climate, examples include: The El Niño Southern Oscillation (ENSO), the Pacific Decadal Oscillation (PDO) and the Madden-Julian Oscillation (MJO) in the tropics, and the North Atlantic Oscillation (NAO) in the extra tropics. Simulating these patterns helps to reveal their teleconnections (long distance links) with regional climate and extreme weather. This underpins capability in initialised climate prediction at multiple timescales, contributing to improved preparedness and mitigation against the socio-economic impacts of variable climate and extremes.

Current activities include:

  • Investigating tropical dynamics and extratropical teleconnections, including the El Nino-Southern Oscillation, La Nina/El Nino symmetries and the teleconnections and impacts it has on initialised climate predictions, especially for China.
  • Grant Award: University of Exeter – Predictability of Regional Climate (to understand the dynamical mechanisms and predictability of East Asian climate, at seasonal and decadal timescales.)
  • Understanding the probability of extreme events in the current climate using the method of Unprecedented Simulated Events using ENsembles (UNSEEN). Application of this to China reveal, for example, that in southeast China there is currently a 10% chance of an unprecedented hot summer month.
  • Developing UK and Chinese seasonal and decadal climate prediction systems, assessing forecast skill and errors and identifying links between model bias and skill.  
  • Grant Award: University of Reading – Aerosols and regional climate dynamics (understanding the influence of aerosol forcing on atmospheric circulation and regional climate, to improve skill of decadal predictions and reduce uncertainty in long-term projections).
  • Building predictive capability by analysing the effects of model resolution and other technical aspects of initialised climate prediction, as well as including predictions from China into the World Meteorological Organization (WMO) Lead Centre’s multi model forecasts.


This work package should contribute to an improved understanding of the mechanisms of climate variability globally, their predictability and effect on China. It also helps to generate new forecast capability, while new climate prediction services should support climate-resilient decision making for socio-economic development. The partnership has already developed a prototype climate service for seasonal forecasting of rainfall in the Yangtze River basin and plans to enhance this capability in coming years. The prototype service is issued by CMA to relevant stakeholders to support robust decision making for flood management.

Scientific highlights

Seasonal forecasts for the Yangtze River Basin. A prototype climate service has been trialled “producing a real-time seasonal forecast for the Yangtze river basin throughout the spring and summer of 2016” providing advice for decision makers at hydroelectric dams. (2017) – See WP5.  View an infographic on this prototype climate service here

Unprecedented events. An innovative technique for assessing current risk of unprecedented climate extreme events UNSEEN (UNprecedented Simulation of Extremes with ENsembles). (2017) – See also WP3 & 5

Li et al 2016 and Thompson et al 2018