Effect of the equatorial Atlantic cold tongue biases on the projected changes of Indochina peninsula summer rainfall in a warmer world from CMIP5/6 models
摘要
Indo-China Peninsula (ICP) rainfall change is an important socio-economic concern. From the historical simulations and future warming scenarios in the Coupled Model Intercomparison Projects, this study finds that climate models tend to project a slight future increase of ICP summer rainfall, but with a large inter-model spread. Moreover, the inter-model spread of sea surface temperature (SST) projections under the North Atlantic intertropical Convergence zone (ITCZ) is identified as an important source of that of the projected ICP rainfall changes. A stronger North Atlantic ITCZ SST warming could significantly enhance the ICP atmospheric convection and strengthen the local rainfall through the pantropical ocean–atmosphere interactions, and vice versa. In particular, the projected uncertainty of North Atlantic ITCZ SST warming is further ascribed to the simulation biases of Atlantic cold tongue (CT) in models. On the one hand, models with a warmer Atlantic CT bias would induce a southward migration of North Atlantic ITCZ during boreal summers and in turn result in a weaker negative cloud-radiation feedback over the observed North Atlantic ITCZ region, favoring local future SST warming. On the other hand, a warmer North Atlantic ITCZ could induce stronger cross-equatorial winds, further increasing the ITCZ SST via the wind-evaporation-SST feedback. As a result, models with a warm Atlantic CT bias would project a stronger future increase in ICP summer rainfall by affecting the North Atlantic ITCZ SST warming. After the corrections with the observed Atlantic CT, the projected ICP summer rainfall increases are significantly decreased, with a reduced inter-model uncertainty. All this highlights the importance of improved equatorial Atlantic CT simulations for the reliable ICP summer rainfall projections.