英文摘要: | A critical aspect of human-induced climate change is how it will affect precipitation around the world. Broadly speaking, warming increases atmospheric moisture holding capacity, intensifies moisture transports and makes sub-tropical dry regions drier and tropical and mid-to-high-latitude wet regions wetter1, 2. Extra-tropical precipitation patterns vary strongly with longitude, however, owing to the control exerted by the storm tracks and quasi-stationary highs and lows or stationary waves. Regional precipitation change will, therefore, also depend on how these aspects of the circulation respond. Current climate models robustly predict a change in the Northern Hemisphere (NH) winter stationary wave field that brings wetting southerlies to the west coast of North America, and drying northerlies to interior southwest North America and the eastern Mediterranean3, 4, 5. Here we show that this change in the meridional wind field is caused by strengthened zonal mean westerlies in the sub-tropical upper troposphere, which alters the character of intermediate-scale stationary waves. Thus, a robust and easily understood model response to global warming is the prime cause of these regional wind changes. However, the majority of models probably overestimate the magnitude of this response because of biases in their climatological representation of the relevant waves, suggesting that winter season wetting of the North American west coast will be notably less than projected by the multi-model mean.
Stationary waves arise from longitudinal asymmetries in topography, diabatic heating and transient eddy heat and vorticity fluxes. The character of the forced waves depends not only on these asymmetric forcings, but also on the zonal mean flow and nonlinear wave–wave interaction, with the additional complication that the asymmetric forcings and zonal mean flow are, in turn, affected by the stationary waves6. In the NH winter, climate models predict that stationary wave changes will form an important component of mid-latitude circulation change7, 8, 9 and past studies have variously attributed these changes to altered wave forcing from the tropics8, 10, 11, 12 or an altered zonal mean basic state in which the stationary wave activity propagates7, 13, 14, with a decisive explanation remaining elusive. Here, we focus on the latest model projections of future eddy meridional wind (v∗), given its importance for regional hydroclimate5. Figure 1 presents an analysis of the Future–Past difference simulated by 35 Coupled Model Intercomparison Project, phase 5 (CMIP5) models (see Methods). The 300 hPa response (Fig. 1b) is dominated by an approximately zonal wavenumber 5 pattern, reminiscent of the circumglobal teleconnection pattern prevalent in natural variability15. It is fairly barotropic (Fig. 1c) and, over North America, the low-level west coast southerlies and interior southwest northerlies contribute to wetting the US west coast and drying the interior southwest5. Such a response has been identified in a number of past studies8, 11, 12 and its structure is robust across the models, but there is a wide spread in magnitude (Fig. 1d). The implications of this spread for North American hydroclimate are seen in Fig. 1e, where the models are divided based on the strength of the 300 hPa interior southwest v∗ anomaly. The stronger half exhibits more west coast wetting and southern drying, as would be expected given the mean flow contributions to this precipitation–evaporation (P–E) change5. The relationship between v∗ and P–E is presented in this format for use in the following analysis, but a similar assessment through correlation between southwest v∗ and P–E exhibits similar features, with a correlation with west coast wetting of up to 0.77 and southwest drying of up to 0.56.
|