Sharing knowledge
© Crown Copyright
Deep convective cloud – cumulonimbus – reaches the tropopause and spreads out. : This link opens in a new window
Deep convective cloud – cumulonimbus – reaches the tropopause and spreads out.

Output 1 - highlight results

Understanding and modelling of African climate and its drivers

To improve systems for climate early warning (e.g. drought or flood) we must first improve our understanding of the factors that drive climate variability and change over Africa and consolidate this new understanding in our main prediction tools - computer models of the climate system. This is the work of Output 1 of the CSRP.

Highlight 1: Improved modelling of African seasonal rainfall

The Met Office Hadley Centres climate model HadGEM3 has undergone two major upgrades during the CSRP programme so far. Assessments of model performance over Africa form part of the model development process and inform priorities for model development.

Key finding:

Upgrades to the HadGEM3 climate model implemented during the CSRP project have improved the model's representation of rainfall in all important sub-Saharan rainy seasons. Substantial improvements (more than 50% reduction of errors found in previous model versions) have been achieved in some regions.

Key impact:

The improved performance of HadGEM3 over Africa has pulled through into improved operational seasonal forecasts, better capability to provide new user-relevant forecasts Output 2 and improved capabilities for greater geographical detail in forecasts Output 3.

Highlight 2: Understanding the West African Monsoon (WAM)

An initial consultation into user needs identified a request for improved predictions of the timing of onset of African rainy seasons. Investigations into how well climate models represent onset has therefore been a key focus of research, starting with analysis of the West African Monsoon season (typically July to September). A key finding so far is that the latest Met Office climate model (HadGEM3) simulates the onset of the WAM over the Sahel much more realistically than models participating in IPCC's Fourth Assessment Report (AR4). This good simulation suggests that key factors controlling onset are well represented in HadGEM3 and that we can use this model as an effective tool to analyse and better these driving processes.

Key finding:

The representation of land surface in the Sahel and the remote influence of sea surface temperature (SST) are important for forecasting the timing of onset. While many models capture the influence of SST to varying degrees, adequate simulation of land-surface processes remains a challenge for all.

Key impact: 

Understanding of this kind is required to help improve climate models and to establish a physical basis for reduced uncertainty in predictions of African climate variability and change.

Further information on model testing – African climate systems African climate systems (PDF, 162 kB)

Further information on Output 2 results.

Highlight 3: Understanding remote influences on African climate rainfall variability

Understanding remote influences on African climate rainfall variability

Year-to-year variations in rainfall over many parts of Africa are influenced by global patterns in sea surface temperature (SST). For example, fluctuations in SST in the equatorial East Pacific that occur with El Niño and La Niña events are known to influence seasonal rainfall over parts of the Greater Horn of Africa during the short-rains (October to December) season. These influences are known as 'teleconnections' because they act from a distance. Because of their large influence on African rainfall it is crucial that climate models are able to simulate these teleconnections. CSRP has conducted the first continent-wide assessment to discover how well a range of climate models represents these important teleconnections. The assessment is being used to identify the underlying model development needed to improve simulation of these remote influences on African rainfall.

Key finding:

The current generation of climate models have inadequate representations of important remote influences (teleconnections) on African rainfall. Model errors in sea surface temperature variability are found to be a key cause of this.

Key impact:

Improved simulation of important rainfall teleconnections is needed to increase confidence in model climate change projections for Africa. These results provide an important guide for targeting research needed to optimise benefit over Africa

Further information on  model testing – Africa teleconnections model testing - Africa teleconnections (PDF, 664 kB)

Further information on Output 3 results

Last Updated: 10 May 2012