Introduction

Background

The current phase of the Coupled Model Intercomparison Project (CMIP6) will have numerous general circulation models (GCMs) to better understand past, present and future climate change (Eyring, Bony et al. 2016). Future projections on climate change are studied and presented under Shared Socio-economic Pathways (SSPs). These are representations of how global society, demographics and economics might change over the next century. The five main SSPs are  SSP1-1.9, SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5 with carbon dioxide (CO2) concentrations ranging from 393 to 1135 ppm for the lowest (SSP1-1.9) and highest (SSP5-8.5) by 2100 respectively (Meinshausen, Nicholls et al. 2020). These projections use a wide range of climate variables that critically contribute to the characterization of Earth’s climate (WMO 2022). With the available models, climatic variables and SSPs, there are tens of thousands of possible model simulations of projected climate change (Eyring, Bony et al. 2016). The sixth Assessment Report (AR6) on Climate Change has already used over 50 GCMs and Earth System Models (ESMs) of CMIP6 to provide a broad context on climate change (IPCC 2021, Masson-Delmotte, Zhai et al. 2021).

However, global outputs are usually too coarse for meaningful contribution to climate change impacts and adaptation. Often, climate simulations are downscaled to regional climate data products (Mahony, Wang et al. 2022, Wang, Fan et al. 2022). There are various approaches to downscaling climate data. One of the suggestions is the use of multi-model ensembles (MMEs) of a minimum of eight to ten simulations to achieve a robust estimate of the mean value of some global and regional quantities (Mote, Allen et al. 2016). Selection of the ensemble should prioritize number of simulation per model, low to moderate bias and moderate to high spatial resolution (Mahony, Wang et al. 2022).

The African continent fosters a strong need for more information and research on regional climate modelling (Shongwe, Pirani et al. 2014, James, Washington et al. 2018). There are no models developed on the continent and the global models previously used have had shortcomings especially with projections on precipitation patterns (Whittleston, Nicholson et al. 2017). Previous regional studies in Africa have confirmed that individual model biases are considerably improved by use of multi-model ensembles (Endris, Omondi et al. 2013, Gbobaniyi, Sarr et al. 2014, Carvalho, Santos et al. 2017, Whittleston, Nicholson et al. 2017). Additionally, it is rather difficult to effect a comprehensive use of all available GCMS with dozens of model parameters and the attached computational costs (Knutti, Furrer et al. 2010).

The purpose of this study is to classify 23 CMIP6 models according to their range of projections and guide regional model selection for an ensemble to project regional or continental climate change over Africa under different climate scenarios. 

Research questions

To achieve my main objective, I ask:

1. What are the similarities and divergences among GCMs to project regional and continental climate change in Africa in the period 2041-2060 under SSP2-4.5?

2. What GCMs project the "best", "middle" and "worst" climate scenarios for Africa and its regions in the period 2041-2060 under SSP2-4.5?