Systematic Climate Model Rainfall Biases over Southern Africa: Links to Moisture Circulation and Topography

Systematic Climate Model Rainfall Biases over Southern Africa: Links to Moisture Circulation and Topography
复制标题

DOI:
10.1175/jcli-d-18-0008.1
复制
发表时间:
2018-08
期刊:
影响因子:
4.9
通讯作者:
C. Munday;R. Washington
C. Munday;R. Washington
中科院分区:
地球科学2区
文献类型:
--
作者:
C. Munday;R. Washington

文献摘要

被引文献

相似文献

气候科学面临的一个重要挑战是了解区域环流和降雨对全球变暖的响应。不幸的是,用于预测未来变化的气候模型很难代表当今的降雨和环流,特别是在区域范围内。南部非洲就是这种情况,来自耦合模式比较项目(CMIP5)第五阶段的模型高估了夏季降雨量,与观测值相比高估了300%,并倾向于低估马达加斯加和西南印度洋的降雨量。在本文中,我们探索与降雨偏差相关的气候过程,目的是评估CMIP5集合的可靠性,并强调模式发展的重要领域。我们发现,在南部非洲湿润的模式中,高降水率与一次异常的东北水汽输送(~10-30g kg−1 S−1)有关,该输送穿过坦桑尼亚和马拉维的高地形,进入副热带南部非洲。这种输送发生在更干燥的模式和再分析数据中所看到的向马达加斯加的东南部转向的优先方向上。我们证明了低层气流中地形相关的模式偏差对于解释降雨中的模式间扩散是重要的;与干燥模式相比,较潮湿的模式对即将到来的东北气流的阻挡倾向较小。不同模式之间低层气流的差异与上游风速和地形的模式表示有关,这两者都是模式发展的重点。
An important challenge for climate science is to understand the regional circulation and rainfall response to global warming. Unfortunately, the climate models used to project future changes struggle to represent present-day rainfall and circulation, especially at a regional scale. This is the case in southern Africa, where models from phase 5 of the Coupled Model Intercomparison Project (CMIP5) overestimate summer rainfall by as much as 300% compared to observations and tend to underestimate rainfall in Madagascar and the southwest Indian Ocean. In this paper, we explore the climate processes associated with the rainfall bias, with the aim of assessing the reliability of the CMIP5 ensemble and highlighting important areas for model development. We find that the high precipitation rates in models that are wet over southern Africa are associated with an anomalous northeasterly moisture transport (~10–30 g kg−1 s−1) that penetrates across the high topography of Tanzania and Malawi and into subtropical southern Africa. This transport occurs in preference to a southeasterly recurvature toward Madagascar that is seen in drier models and reanalysis data. We demonstrate that topographically related model biases in low-level flow are important for explaining the intermodel spread in rainfall; wetter models have a reduced tendency to block the oncoming northeasterly flow compared to dry models. The differences in low-level flow among models are related to upstream wind speed and model representation of topography, both of which should be foci for model development.