The future of the East African Long-rains: insights from observations, model experiments, and process-based model assessments

East Africa's two rainy seasons, known as the Long (March-May) and Short (October-December) rains, are the lifeblood of the region. Their timely arrival is vital for all aspects of the region's socio-economic system, from rainfed agriculture and hydroelectric power generation to aquifer recharge and flood management. However, the timing and intensity of the rainy seasons varies significantly on interannual and interdecadal timescales, causing extreme droughts and floods (Palmer et al, 2023). Whilst this variability is characteristic of the East African climate, there are concerns that anthropogenic forcings may be modulating this variability. Although there is a high degree of confidence that the Short-rains are becoming wetter (Wainwright et al, 2019), the future of the Long-rains remains uncertain: satellite data indicate a recent drying trend, whereas models from the Coupled Model Intercomparison Project (CMIP) generally project increased rainfall, albeit with significant variations between the models (Cook et al, 2020). This paradox creates uncertainty, which makes it difficult for stakeholders to implement adaptation strategies (Rowell et al 2015).

One approach to narrowing this uncertainty is by constraining climate model projections through process-based model assessments (James et al, 2018). Rather than assessing the performance of a model based on its ability to replicate historic climate statistics, process-based assessments investigate the meteorological dynamics and physical processes simulated by climate models. For example, hindcasts can be compared to reanalysis data to judge whether models capture the mechanisms driving historic climate variability, or future simulations can be interrogated to assess the plausibility of mechanisms responsible for projected future changes (James et al, 2018). The success of a process-based assessment depends on identifying the key mechanisms controlling regional climate, yet no single mechanism can fully explain the observed variations in the East African Long-rains (Palmer et al, 2023). Rather, numerous potential mechanisms have been identified, including the effects of the Madden-Jullian-Oscillation (Pohl and Camberlin, 2006), westerly wind anomalies from the Congo Basin and the influence of Indian Ocean Tropical Cyclones (Finney et al, 2020), large-scale shifts in the latitudinal migration of the tropical rainbelt (Wainwright et al, 2019), and variability in the Indian Ocean Walker Cell (Funk et al, 2023.) Many of these mechanisms are thought to be related to SST patterns (Funk et al, 2023). Process-based assessments are therefore currently unable to decisively constrain projected changes in the Long-rains (King et al, 2021).

Aims

The aim of this study is to improve our understanding of the mechanisms controlling the long-rains through observational data and model experiments, and then to apply this new understanding to constrain projected changes in the long-rains through process-based assessment of CMIP models.