Delivering a stepchange in climate science for Africa
FCFA has made significant progress in improving understanding and modelling of Africa’s climate and applying this to real world problems. While there are large gaps in our understanding of the climate of Africa and future projections are still uncertain, FCFA has contributed towards a step change in climate science for the continent. This section highlights FCFA’s key climate science advances and scientific outputs.
Delivering new model developments
A key objective of FCFA’s IMPALA project was to develop a step change in the capabilities of climate models to simulate Africa’s climate. Model developments from the FCFA programme have been key to improving the representation of climate processes across Africa in climate models.
Incorporating advanced understanding of Africa’s climate in the global MetUM model.
Improvements in understanding key processes of Africa’s climate, driven both locally and remotely, from the FCFA programme have been incorporated into the 7th iteration of the Global Atmosphere (GA7) Met Office Unified Model (MetUM). This has allowed for the global models to more accurately represent Africa’s climate through the new and improved understanding emerging from FCFA. Further improvements from the programme are also being incorporated into the 8th iteration of the model MetUM GA8.
Delivering the first Pan-African convection-permitting model
FCFA, through IMPALA, has delivered the first Pan-African high-resolution (4.5km) model called CP4-Africa ( Stratton et al., 2018 ). This model is not only able to more accurately represent convective systems, rainfall intensities and the diurnal cycle, but also in many cases large scale circulation over Africa. The improvement in simulating convection is not only beneficial in projection of rainfall extremes, dry spells and high winds, but also continental-scale circulation and regional rainfall (Senior et al. 2021, which synthesises a large number of science papers from across all the FCFA projects).
The programme also produced a guide to accessing and using CP4-Africa data ( Senior et al. 2020 ), which supports users to use the new data with some confidence - understanding limitations and new information in the context of existing model data.
Model evaluation and developing diagnostic metrics
Building on FCFA’s advances in understanding local processes, the Model Evaluation Hub was developed to draw on local expertise to perform targeted process-based evaluation for important climate processes over the continent ( James et al. 2018 ). The first phase of the hub, LaunchPAD, has focused on developing region-specific tools for model evaluation in Africa, through key partnerships between UK and African scientists. Research fellows from LaunchPAD have been collaborating with software developers in order to develop key diagnostic tools, which will be tested by the Met Office before being made publicly available.
“Having ring fenced activities for model development meant that we were able to see more established climate science into impact. Right from the model development to actually delivering models and seeing impact on the ground in the four year time frame, which is almost unprecedented”
(FCFA climate scientist reflecting on progress made during a learning moment in 2020.)
peer reviewed articles, working papers and policy briefs published on issues of climate science and its application in Africa.
Understanding past and future climate extremes over Africa
A key component of FCFA’s research focused on how climate change has been and will continue to have an impact on climate extremes across Africa.
Improved understanding of convection and rainfall through new convection-permitting models
Using CP4-Africa simulations, FCFA has delivered a new understanding of convection (i.e. the clouds and thunderstorms that generate most rainfall in the tropics) providing a better representation of rainfall extremes over the continent. As the first pan-African convection-permitting climate model, CP4-Africa represents small-scale processes such as storms, which have impacts on wider scales, as well as local climate processes and so it can provide improved information, including information at a finer scale than previously possible.
The analysis of CP4-Africa has highlighted considerable new information and improved capability, but also exposed remaining biases. CP4-Africa is also only driven by a single global model, therefore, for decision making, it should always be considered together with information from other contemporary models, such as those from CMIP and CORDEX.
The model suggested that extreme rainfall events in the Sahel and East Africa, which occur once every 30 years (rainfall exceeding 60mm rainfall in 3 hours over 25kmx25km area), may happen as often as every 3-4 years by the end of the century, under continued greenhouse gas concentrations represented by the RCP8.5 scenario . While dry spells during wet seasons over Western and central Africa are also projected to happen twice as frequently in the future compared to the present-day ( Kendon et al, 2019 ). The CP4-Africa model has also been used to project future changes in lightning across the continent, with results indicating that total days with lightning may decrease but those days with lightning will be more extreme (i.e. have more flashes) ( Finney et al. 2020 ).
CP4-Africa data was also able to project regional changes across the continent: In East Africa , the models projected increases in annual rainfall amounts (±40%) and increasing extreme rainfall (±50%) by the end of the century ( Finney et al 2020 ). In West Africa , CP4-Africa simulations provided new insight into how megastorms in the region may change, with rainfall becoming increasingly intense and dry spells becoming longer ( Berthou et al.,2019 ). CP4-Africa data was also used (in parallel with a parameterised simulation at 25 km grid spacing; P25 climate model) to compare the present climate of West Africa with the projected climate at the end of the century. These projections indicate that typical future temperatures will be much higher than current temperatures and typical rainfall events at the end of the century will be as intense as the current 1 in 20 rainfall event ( Fitzpatrick, 2020 ).
Improving the understanding of trends in megastorms in the Sahel
Using satellite observations from the past 35 years, AMMA-2050 has furthered understanding of how megastorms or Mesoscale Convective Systems (MCSs) over the Sahel region are responding to climate change ( Taylor et al, 2017 ). In this region global warming has increased temperatures in the Sahara desert, which has been linked to the increasing frequency of intense megastorms in the West African Sahel. Megastorms have tripled in frequency since the 1980’s due to climate change. Research has also demonstrated how the increasing frequency of intense storms is contributing to the sharp rise in reported flood events, alongside increasing urbanisation. Projections for future rainfall in the region suggest a 28% increase in extreme rainfall associated with MCSs ( Fitzpatrick et al. 2020 ).
Research has also shown how daily rainfall has changed in southern West Africa, with less frequent but more intense rainfall during the first rain season (April - July) over the last 2 decades and more frequent and intense rainfall during the second season (September - November) over the last decade ( Nkrumah et al. 2019 ). Examining past trends over countries in the Sahel, show that rainfall in the central and western Sahel has increased due to more rainy days, with longer wet spells and shorter dry spells. While along the Guinea Coast, slight increases are due to more intense rainfall with shorter wet spells ( Sacré Regis et al., 2020 ).
Understanding the processes that influence Africa's climate
A key component of FCFA’s climate science work included improvement in understanding and modelling the key processes that influence local climate variability and change in Africa. This included investigating the processes that have driven recent changes and understanding the influence of local (e.g. land-atmosphere processes) and remote climate processes (via the "teleconnections").
Understanding what drives the variability and change of the East African long and short rains
FCFA has made significant progress in understanding the East African short and long rainy seasons. In terms of the long rains, Research from IMPALA has improved understanding of the processes which produce between 30-50% of interannual rainfall variability in the region ( Vellinga and Milton, 2018 ). Further research from IMPALA and HYCRISTAL has also shown that some of the Indian Ocean’s influence of rainfall over the region (e.g. Indian Ocean Dipole ) are poorly represented in models, which influences how models represent both the short rains ( Hirons and Turner, 2018 ) and long rains (Sabiiti, et al. in prep.).
HyCRISTAL has also made significant progress in understanding the East African climate paradox where climate projections suggest long-rains rainfall will likely increase but recent observations have shown rainfall has declined. Research from FCFA has identified a recent recovery of the long rains and been able to attribute the observed rainfall decline over the region to changes in the long rains which have started later and ended earlier, linked to changes in the Arabian heat low and sea surface temperatures ( Wainwright et al., 2019 ). New insight has also shown how changes to the Sahara Heat Low in the future may result in particular increased rainfall during the short rains, with the short rains starting later but lasting longer ( Dunning et al, 2018 ). Further research has also shown the influence of zonal winds across the Congo basin and Gulf of Guinea in influencing variability of the long rains, including the paradox drying ( Finney et al., 2019b; Walker et al. 2020 ).
Understanding storms over the East African highlands, coast Lake Victoria
HyCRISTAL has also improved understanding of storm formation over Lake Victoria. The improved representation of convection in models, has provided researchers with a better picture of how storms over the Lake Victoria Basin develop ( Woodhams et al., 2019 ). CP4-Africa simulations have also led to improved understanding of the intensity and diurnal cycle of rainfall over the coast and Lake Victoria Basin ( Finney et al., 2019a ), as well as the Ethiopian Highlands and South Sudan (Misiani et al., 2021).
Understanding how atmospheric and oceanic processes impact rainfall over East Africa
FCFA has improved the understanding of the role of various atmospheric and oceanic features that control rainfall variability over East Africa. Research from IMPALA and HYCRISTAL has shown roughly 15% of day-to-day variability in rainfall over East Africa can be explained by large-scale waves in the atmosphere, known as the Kelvin waves ( Jackson et al., 2019 ). Research from HyCRISTAL has demonstrated the role of the Indian Ocean Dipole and sea-surface temperatures in the western Indian Ocean contributed to the extremely wet short rain seasons in 2019 ( Wainwright et al. 2020 ) and so the record-breaking levels of Lake Victoria and associated floods. Other atmospheric oscillations such as the Madden-Julian Oscillation (MJO) and the presence of tropical cyclones in the Indian Ocean has also been shown to lead to moisture flow from Congo to East Africa and enhanced rainfall in the region ( Finney et al., 2019b ), with both expected to change under climate change. Research has also indicated that the coastal region of East Africa may not experience the same increase in rainfall that is expected inland, as a result in changes in sea-breezes ( Finney et al., 2019c ).
Understanding the influences on megastorms in the Sahel
FCFA has improved the understanding of processes and drivers which influence megastorms or Mesoscale Convective Systems (MCSs) over West Africa. New research from AMMA-2050 has made advances in tracking these storms by linking land surface conditions with convection. This research has shown how dry soils can produce more convection activity linked to megastorms and have a significant influence on the path that these storms follow ( Klein and Taylor, 2020 ).
New understanding has also been developed in terms of how future thermodynamics and storm dynamics affect extreme rainfall associated with MCSs ( Fitzpatrick et al. 2020 ). Further research has also shown that wind fields have a significant influence on these MCSs, and this influence is not well represented in conventional climate models ( Bickle et al. 2020 ). Research has also been extended to offer new insight into changing seasonality and drivers of MCSs over south west Africa ( Klein et al. 2020 ).
Research into the annual cycle of rainfall over West Africa, have provided new insights into how misrepresentations of the position of the ITCZ in CMIP affect the timing and amount of rainfall which is simulated over West Africa. ( Sow et al. 2019 ).
Understanding how changes to warming oceans influence rainfall over the Sahel
AMMA-2050 has also highlighted how large-scale and regional-scale atmospheric responses to increases in sea-surface temperature may impact rainfall over the Sahel region ( Dixon et al., 2019 ). Research has also shown how sea-surface temperatures in the Mediterranean can be used as a predictor for heavy rainfall over the Sahel region ( Diakhaté et al. 2020 ).
Identifying the emergence of new climate regimes in West Africa
Research from AMMA-2050 has identified when the impacts of climate change will clearly emerge above the strong natural year-to-year and decade-to-decade variations in rainfall which characterise the region. This research has shown that the western Sahel may begin experiencing a much drier climate between 2028 - 2052, while the eastern Sahel could begin to experience a much wetter climate before 2054 ( Gaetani et al. 2020 ).
Understanding the link between the climate of central and southern Africa
UMFULA has provided new knowledge on the links between the climates of central and southern Africa. Research has shown that the African Easterly Jet South, which is a key driver of convection over central Africa, is regulated by mid-level high pressure systems in the Kalahari region and southern subtropical westerly waves ( Kuete et al. 2020 ). Research has also examined how the Congo Air Boundary and Kalahari Drylines play important roles in variability of spring rainfall for Southern Africa ( Howard & Washington, 2019 ). New insights have also been gained into how night-time easterly low level jets transport moisture from the Indian Ocean through valleys in the Great Rift system to the Congo Basin, with stronger jets being associated with drought in Southern and Eastern Africa ( Munday et al 2020 ).
Improved understanding on convection and cloud formation
Through the use of the CP4-Africa simulation over Southern Africa, FCFA has improved understanding of the seasonal cycles of Tropical Temperate Cloud Bands (TTCBs). Research from UMFULA has indicated a decrease in TTCBs with climate change, which is linked to projected future declines in rainfall over Southern Africa ( Hart et al. 2018 ). Research has also examined the importance of Sea-Surface Temperatures (SST) on atmospheric circulation and the formation of tropical cloud bands over the region ( Desboilles et al. 2018 ). FRACTAL has also demonstrated methods for characterising the tropical rainbelt (according to intensity, location and width) which proved beneficial for assessing models and projections of tropical rainfall over Southern Africa. ( Nikulin et al. 2019 ).
Understanding how key climate processes in central and southern Africa interact with ENSO
The El-Nino Southern Oscillation (ENSO) is a significant driver for interannual rainfall variability across South Africa. FCFA has provided new insights into how other key processes influencing rainfall interact with ENSO. Research from UMFULA has shown how regional circulations such as the weakened Angola low and South Indian highs, and strengthened Botswana High played a role in exacerbating El Nino- related droughts of 2015/2016 ( Blamey et al. 2018 ). Research on the Angola low , has also identified 2 phases (the Angola heat low and tropical low) which has improved understanding of rainfall variability and the interaction with ENSO in the region ( Howard and Washington, 2018 ). Progress has also been made in understanding tropical lows (similar to very weak cyclones forming in eastern Angola) and the influence of El Nino in reducing the number of lows forming, leading to reduced rainfall in many Southern African countries ( Howard et al. 2019 ).
Research from FRACTAL has also furthered understanding of the co-behaviours between ENSO, the Antarctic Oscillation and the Intertropical Convergence Zone (ITZC), particularly in the role of an Antarctic Oscillation in suppressing or enhancing ENSO over Southern Africa ( Quagraine et al. 2019 ). An example of this co-behaviour can manifest as in a positive Antarctic Oscillation in summer enhancing drought producing conditions of the El-Nino cycle (which is poorly represented in CMIP models) ( Quagraine et al. 2020 ).
Understanding how topography affects moisture flows in Southern Africa
UMFULA research has been able to improve the understanding of how topography features influence rainfall over Southern Africa. This included examining how overestimation of rainfall over Southern Africa in CMIP models is partly due to model’s which do not accurately represent the role of topography in regional circulation ( Munday & Washington, 2018 ). Research has also examined how models incorrectly represent a flatter topography over Madagascar leading to simulations of more rainfall than what is actually observed over the region from Mozambique to Angola ( Barimalala et al. 2018 ).
Understanding tropical cyclones in the South West Indian Ocean
FRACTAL has also provided new insights into the paths of tropical cyclones in the South West Indian Ocean. This includes examining how well the paths of cyclones are represented in climate models, which indicated that models overestimate the amount of cyclones forming in the Mozambique Channel and poorly represent interannual variability ( Maoyi et al. 2017 ). Research has also provided new insights into how cyclones may change in the future, with the number of tropical cyclones in the region expected to decrease after 2°C of global warming ( Muthige et al. 2018 ).
Understanding drought in the Western Cape of South Africa
FCFA’s research has contributed to growing knowledge on the drivers of the highly publicized ‘day zero’ drought experienced by the city of Cape Town in the Western Cape of South Africa. Research from UMFULA has been able to partly attribute drought conditions to the poleward shift of moisture corridors and the displacement of jet streams and storm tracks in the South Atlantic ( Sousa et al. 2018 ). The unprecedented drought has also been linked to the expansion of the Hadley cell , and the trend of declining number of rainfall days and a recent decline in rainfall intensity over the region ( Burls et al., 2019 ).
Further research has also been able to link sources of moisture over the west coast of South Africa, to low level jets from South America which feed moisture to key basins in the Atlantic Ocean ( Ramos et al. 2018 ).
Understanding global model uncertainty over Africa
While significant uncertainties still exist in understanding Africa’s future climate, FCFA made progress towards understanding some of the poorly represented processes and uncertainties within global climate models. CP4-A runs demonstrate that we expect the change in rainfall extremes to be underestimated in all global climate models, and there is a need to synthesise uncertainty from global models, with knowledge of such systematic errors across models. Mittal et al (in review) provides one first approach to do this, in the context of metrics for tea.
Investigating uncertainties in future rainfall over East Africa
FCFA aimed to reduce uncertainties in projections for future rainfall over East Africa, through a process of identifying models that misrepresent or do not account for future-relevant processes over the region. Through understanding why different models produce widely different projections for the East African long rains season, HyCRISTAL has been able to reduce the plausible range of projected changes by a third for this season . This is by eliminating models that do not accurately represent key climate processes ( Rowell and Chadwick, 2018 ; Rowell, 2019 ).
Advances have also been made in terms of quantifying the skill of seasonal predictions over the region, providing a further understanding of model errors ( Walker et al. 2019 ). Evaluation of regional climate models (RCMs) to simulate the rainfall over Uganda within CORDEX , has provided new insight into how these models misrepresent the climate of Uganda. This research suggests models present less rainfall than what is currently observed, and while they represent the variability of the dry seasons they inadequately represent variability during the two rainy seasons ( Kisembe et al. 2019 ).
Research has also shown that CMIP5 RCP scenarios do not sufficiently explore the range of plausible air quality policies and their remote influences. Using emission scenarios that account for the range of possible aerosol emissions show that significant emission reductions may result in the later start and end of the short rains in Africa ( Scannell et al. 2019 ).
Investigating future uncertainties over West Africa
FCFA has made advances in understanding rainfall uncertainty across West Africa. AMMA-2050 has provided new analysis of CMIP models, furthering the understanding of future rainfall uncertainty, as a result of poorly represented annual rainfall cycles over West Africa. Improved understanding of uncertainty of future temperature projections , have also shown that CMIP models may underestimate the impacts of regional warming ( Macadam et al. 2020 ).
Research has also improved understanding of uncertainty in future rainfall intensity of the Sahel . This has indicated that uncertainty is produced through intermodel variability in: remote warming, the strength of a narrow band of warming–advection–circulation feedback along the southern Sahara and to some extent the effect of southern Saharan evaporative anomalies, all of which impact lower-tropospheric temperature gradients and hence Sahelian rain storms ( Rowell et al. 2021 ).
“FCFA has demonstrated that with investment in climate science, and quite modest investment in model development, significant progress is possible against “grand challenges”.
(FCFA Project closure report, 2021).
Investigating uncertainties in rainfall over Central and Southern Africa
FCFA has made progress in understanding uncertainty in future rainfall projections over the Congo Basin . Research from UMFULA into future projections for rainfall over the basin, has shown that future changes may be related to the weakening of the Indian Ocean Walker circulation and convection over the Maritime Continent ( Creese et al. 2019 ). Research has also been able to eliminate the largest increases and decreases in the September - November rainfall seasons from CMIP models. This includes furthering the understanding of how rainfall varies between the western and eastern regions of this basin , and demonstrating that the largest increases in the western region are unlikely ( Creese & Washington, 2018 ). As well as improving understanding of how historical wet or dry models affect projections for the eastern region of the basin, indicating that the driest changes are unlikely ( Creese et al. 2019 ).
UMFULA has also furthered understanding of projections for future drying over Southern Africa , indicating that models overestimate current rainfall over the region, and eliminating projections for the largest drying over Southern Africa ( Munday & Washington, 2019 ). Research has furthered understanding of the variability in model representation of tropical-extratropical cloud bands and rainfall over Southern Africa ( James et al. 2020 ).
FRACTAL has produced a new understanding of uncertainties through comparisons of Regional Climate Models (RCMs) and Global Climate Models (GCMs) . This research has shown that although there is uncertainty over the scale of changes, GCM and RCM project a similar future ( Dossio et al. 2018 ). Changes in precipitation and circulation have also been examined for Southern Africa, demonstrating how contradiction between GCMs and RCMs are a result of inconsistencies in the physical parameterizations of precipitation rather than inconsistencies in regional circulation patterns ( Pinto et al. 2018 ). Research has also shown how the formulation of RCM plays a significant role in spatial bias of seasonal rainfall, while the magnitude of the bias is influenced by model resolution ( Wu et al. 2020 ).
Understanding the impacts of climate change in Africa
One of the key ambitions of FCFA was to improve the use and uptake of climate information in medium- to long-term decision making. In addition to improving the fundamental understanding of Africa’s climate and how it might change, research also focused on how future changes may impact on key sectors across the continent.
FCFA has provided new insights into how climate change may impact agriculture across the continent. Research using global CMIP and Cordex models along with CP4-Africa simulations have been able to demonstrate how rising temperatures and rainfall declines in certain regions may lead to declines in key staple crops such as maize, soybean and cassava ( Chapman et al. 2020 ).
The AMMA-2050 team has also provided further insights into crop losses in West Africa, showing how climate change has already led to reduced yields for sorghum and millet ( Sultan et al. 2019 ). Research has also shown how the breeding of new maize crops needs to consider how the climate will change within the breeding and adoption period to avoid future losses ( Challinor et al. 2016 ). Research has also helped to identify the most suitable varieties of millet in West Africa given expected climate change ( Rhone et al. 2020 ).
The joint HyCRISTAL and UMFULA Climate Information for Resilient Tea Production (CI4Tea) project has also been able to provide insights into how the tea sectors of Kenya and Malawi may be impacted by climate change (Mittal et al. In press).
FCFA has improved the understanding of how climate change may affect water resources across the continent.
UMFULA has provided new knowledge on how climate change may affect future water availability, and provided an evidence base for decision-making within the Water-Energy-Food-Environment (WEFE) nexus. This includes investigations of rainfall over river basins with key socio-economic importance, such as Mbarali sub-catchment in the Rufiji River Basin ( Mutayoba et al. 2018 ), and the Limpopo River Basin ( Rapolaki et al. 2019 ). Further research includes assessments of how global hydrological models can be regionalised, by applying these to the Rufiji River Basin ( Siderius et al. 2018 ).
UMFULA has also provided new evidence of the hydrological response of ENSO in countries in Southern and Eastern Africa ( Siderius et al 2018 ), and linking water supply disruption in Botswana to the 2015/2016 El Niño event ( Gannon et al. 2018 ). Further research has also demonstrated how decreased rainfall during the 2015/2016 El Niño event led to reduced groundwater recharge and water levels ( Kolusu et al. 2019 ).
HyCRISTAL has also provided insights in water resource management in the Lake Victoria Basin. The use of CP4-A models has provided new insights into extreme rainfall over the basin, helping understand how river flows may change in the future . HyCRISTAL has analysed the possible role of climate change in recent record breaking Lake Victoria lake levels ( Wainwright et al. 2020 ) and has provided plausible future lake levels and outflows for Lake Victoria to inform planning for transport infrastructure and management through the HyCRISTAL Transport Pilot Project (HyTpp).
In West Africa, AMMA-2050 has provided new insights into how river flows in the Senegal and Niger river have changed since mid-twentieth century ( Wilcox et al. 2018 ). Research has also contributed to understanding how climate, land-use and land cover change affects river flows and water availability in South-Western African Basins ( Obahoundje et al. 2018 ) and the Bandama Basin in Côte D’Ivoire ( Kouame et al. 2019 ).
Research from FCFA has also demonstrated the potential impact of climate change on energy generation. UMFULA has provided an evidence base for balancing hydropower generation, with water, irrigation and environmental demands in the Shire River Basin and Rufiji River Basin . Research has furthered understanding of climate risks for hydropower plans in southern and eastern Africa ( Conway et al. 2017 ). UMFULA has also demonstrated the link between power supply disruptions and drought associated with the 2015/2016 El Niño event ( Gannon et al. 2018 ).
AMMA-2050 has provided research which demonstrated how climate change may change the potential for solar power ( Bichet et al. 2019 ). Research has also examined the impacts on hydropower production across West Africa ( Abohoudje et al. 2017 ), within the Bandama Basin in Côte D’Ivoire ( Kouame et al. 2019 ).