Climate models are vital to the assessment of the impacts of climate change in the Central African regions. Establishing how well models reproduce key processes is important to the confidence we attach to these tools. This study examines model representation of the September to November characteristics, such as location and intensity, of the African Easterly Jet (AEJ) north and south in a sample of 16 commonly evaluated CMIP5 and CMIP6 models and in two reanalyses (ERA5 and MERRA2). The analysis evolves to assess key drivers of the AEJ from energetic interactions, the characteristics of mid-level highs and thermal lows and the nature of surface thermal heating. Over West Africa, several models miss the southeast-northwest orientation of the AEJ-N core, leading to a gap of around 6⁰ in the location of the jet while most CMIP5 models also fail to locate AEJ-S over southern Central Africa. In general, the spread of simulated AEJ locations around reanalyses is larger for the CMIP5 sample compared to CMIP6 equivalent models, indicating improvement from CMIP5 to CMIP6 in this 16 model subset. However, this improvement in some CMIP6 models (e.g. GISS-E2-1-G and MIROC6) is not related to a maximum surface meridional gradient in temperature. Most CMIP5 and CMIP6 models underestimate the surface temperature gradient over AEJ-N region. As a first order diagnostic of the jet's acceleration, most coupled models better simulate the atmospheric energetic interactions over AEJ-N region that leads to its strong contribution to AEJ-N maintenance compared to AEJ-S. This study strengthens our understanding of the mid-level circulation over Central Africa by detecting gaps in the mechanisms maintaining the AEJ in coupled models and highlights processes that should be improved in future ensembles.

气候模型对于评估中非地区气候变化的影响至关重要。确定模型再现关键过程的能力对于我们对这些工具的信心非常重要。本研究在 16 个常用评估的 CMIP5 和 CMIP6 模型样本和两个再分析（ERA5 和 MERRA2）样本中检验了 9 月至 11 月特征的模型表示，例如位置和强度，非洲东部急流 (AEJ) 北部和南部。该分析逐渐从能量相互作用、中层高压和热低压的特征以及地表热加热的性质评估 AEJ 的关键驱动因素。在西非，几个模型错过了 AEJ-N 核心的东南西北方向，导致大约 6 ^0的差距在喷气式飞机的位置，而大多数 CMIP5 型号也未能在中非南部上空定位 AEJ-S。一般来说，与 CMIP6 等效模型相比，CMIP5 样本在再分析周围的模拟 AEJ 位置分布更大，表明在这 16 个模型子集中从 CMIP5 到 CMIP6 的改进。然而，某些 CMIP6 模型（例如 GISS-E2-1-G 和 MIROC6）的这种改进与最大表面经向温度梯度无关。大多数 CMIP5 和 CMIP6 模型都低估了 AEJ-N 区域的表面温度梯度。作为喷气机加速度的一阶诊断，大多数耦合模型更好地模拟了 AEJ-N 区域上的大气能量相互作用，与 AEJ-S 相比，这导致其对 AEJ-N 维护的巨大贡献。