Climate narratives to characterise and communicate uncertainty in climate change in South-eastern Africa

Abstract:

Knowledge about regional and local climate change can inform climate risk assessments and adaptation decisions. But climate model projections indicate deep uncertainty in both sign and magnitude of future precipitation change in South-eastern Africa. Further, multiple climate model projections from multiple generations (CMIP3, CMIP5, CMIP6), makes climate risk assessment challenging for water, energy and agriculture sectors. Malawi faces risks to >90% of electricity generation (hydropower) and irrigation infrastructure due to uncertain future precipitation changes and consequent impacts on Lake Malawi levels. Two climatological regions affect precipitation in the Lake Malawi Shire River Basin: the East African Monsoon Region in the northern half and the Southern African Region in the southern half of the basin. In this study we develop qualitative narratives of plausible future changes in precipitation and associated uncertainty using a structured expert elicitation with regional climate and meteorology experts. We conduct elicitations where experts interpret theory, observations and climate model results to identify key climate processes and how their future changes influence precipitation in the two climatological regions, at long-term (decadal to multi-decadal) and short-term (seasonal to interannual) timescales. Using the most influential climate processes (drivers), we construct narratives of future precipitation change and characterise each narrative by specifying the processes that could influence such a change. Experts identify three subsets of climate processes; influencing East African Monsoon Region, influencing Southern African Region, influencing both. These lead to different drivers and corresponding narratives for the two parts of the Lake Malawi Shire River Basin. Considering that uncertainty in future precipitation change is not well-captured by climate models, especially in data-scarce regions, such narratives enable the characterisation and communication of uncertainty.