9.8.1 Uncertainties, confidence levels and unknowns

• While climate models are generally consistent regarding the direction of warming in Africa, projected changes in precipitation are less consistent.
• The role of land-use and land-cover change (i.e., land architecture in various guises) emerges as a key theme. The links between land-use changes, climate stress and possible feedbacks are not yet clearly understood.
• The contribution of climate to food insecurity in Africa is still not fully understood, particularly the role of other multiple stresses that enhance impacts of droughts and floods and possible future climate change. While drought may affect production in some years, climate variability alone does not explain the limits of food production in Africa. Better models and methods to improve understanding of multiple stresses, particularly at a range of scales, e.g., global, regional and local, and including the role of climate change and variability, are therefore required.
• Several areas of debate and contention, some shown here, also exist, with particular reference to health, the water sector and certain ecosystem responses, e.g., in mountain environments. More research on such areas is clearly needed.
• Impacts in the water sector, while addressed by global- and regional-scale model assessments, are still relatively poorly researched, particularly for local assessments and for groundwater impacts. Detailed ‘systems’ assessments, including hydrological systems assessments, also need to be expanded upon.
• Several of the impacts and vulnerabilities presented here derived from global models do not currently resolve local-level changes and impacts. Developing and improving regional and local-level climate models and scenarios could improve the confidence attached to the various projections.
• Local-scale assessments of various sorts, including adaptation studies, are still focused on understanding current vulnerabilities and adaptation strategies. Few comprehensive, comparable studies are available within regions, particularly those focusing on future options and pathways for adaptation.
• Finally, there is still much uncertainty in assessing the role of climate change in complex systems that are shaped by interacting multiple stressors. Preliminary investigations give some indications of these interactions, but further analysis is required.