Tightening of tropical ascent and high clouds key to precipitation change in a warmer climate

Abstract

The change of global-mean precipitation under global warming and interannual variability is predominantly controlled by the change of atmospheric longwave radiative cooling. Here we show that tightening of the ascending branch of the Hadley Circulation coupled with a decrease in tropical high cloud fraction is key in modulating precipitation response to surface warming. The magnitude of high cloud shrinkage is a primary contributor to the intermodel spread in the changes of tropical-mean outgoing longwave radiation (OLR) and global-mean precipitation per unit surface warming (dP/dT_s) for both interannual variability and global warming. Compared to observations, most Coupled Model Inter-comparison Project Phase 5 models underestimate the rates of interannual tropical-mean dOLR/dT_s and global-mean dP/dT_s, consistent with the muted tropical high cloud shrinkage. We find that the five models that agree with the observation-based interannual dP/dT_s all predict dP/dT_s under global warming higher than the ensemble mean dP/dT_s from the ∼20 models analysed in this study.

Figure 1. A schematic of model simulated changes in the Hadley Circulation and tropical clouds along with the OLR and precipitation changes in a warmer climate.

Black arrows mark the climatological Hadley Circulation with ascents near the equator and descents in the subtropics. Blue triangles indicate the poleward expansion of the descent zone and equatorward contraction of the ascending branch of the Hadley Circulation under global warming. The blue upward arrows indicate the strengthening of ascending motion over the equatorial tropics, where precipitation increases. The light and dark grey shadings of clouds correspond to the mean cloud distributions in the present and future climate, respectively, highlighting the rise of cloud top and the decrease of high cloud cover over the equatorial tropics and the reduction of low cloud amount in the subtropics. The higher cloud top causes decreased OLR over the equator, while the decrease of high cloud cover leads to increased OLR away from the equator. The schematic is based on multimodel-mean climate model simulations in response to increasing CO₂ (refs 29, 30).

Figure 2. Relationship between tropical circulation and high CF sensitivities to surface warming.

(a) Interannual and (b) centennial multimodel-mean zonal-mean high CF (solid curves), precipitation (dashed curves) and vertical pressure velocity at 250 hPa (ω₂₅₀, signed negative for ascending motion, dotted curves) changes per degree of tropical-mean (20°S–20°N) surface temperature increase. The multiyear-mean ω₂₅₀ is shown in black solid curves. S and W indicate the strengthening and weakening segments of the Hadley ascent, and THA marks the tightening of Hadley ascent, defined by the weakening of upward motion at the flanks of the intensifying equatorial ascent. (c) Interannual and (d) centennial tropical-mean high CF change per unit surface warming dCF/dT_s scattered against the change of the tropical ascending area fraction per unit surface warming dF_ω/dT_s for 21 CMIP5 models. The tropical ascending area is defined by ω₂₅₀<0 Pa s⁻¹. Each model is represented by a lowercase letter. Multimodel means are marked in solid coloured circles. The least-squares linear regression lines and correlation coefficients between the x-axis and y-axis variables are shown.

Figure 3. Relationship between tropical high CF and OLR sensitivities to surface warming.

(a) Interannual and (b) centennial tropical-mean OLR change per unit surface warming dOLR/dT_s scattered against tropical-mean high CF change per unit surface warming dCF/dT_s for 21 CMIP5 models. Each model is represented by a lowercase letter. Multimodel means are marked in solid coloured circles. The least-squares linear regression lines and correlation coefficients between the x-axis and y-axis variables are shown. The observed dCF/dT_s from multiple satellite sensors along with the observed dOLR/dT_s from CERES EBAF are shown in black symbols in a. The grey-shaded area marks the uncertainties of the observed data, based on 95% confidence interval of the regression slope between deseasonalized CERES EBAF OLR and HadCRUT4 T_s and the range of the observed dCF/dT_s from the satellite data used.

Figure 4. Emergent constraint on the hydrological sensitivity.

(a) Interannual global-mean precipitation change per unit surface warming L_vdP/dT_s scattered against interannual tropical-mean OLR change per unit surface warming dOLR/dT_s for 23 CMIP5 models. The CERES dOLR/dT_s (the black dot on the x-axis) with the 95% confidence level marked horizontally in grey shading. The GPCP global-mean L_vdP/dT_s is shown in open black circle with the 95% confidence level marked vertically in grey shading. The CERES OLR-constrained L_vdP/dT_s estimate based on the linear regression relation between the interannual L_vdP/dT_s and dOLR/dT_s is marked by grey horizontal lines with their length corresponding to the 95% confidence level. The overlapped range of the GPCP and CERES OLR-constrained L_vdP/dT_s is used as the best estimate of observational interannual L_vdP/dT_s with the 95% confidence level. (b) The temperature-mediated global-mean L_vdP/dT_s scattered against the interannual L_vdP/dT_s for 21 CMIP5 models. The best estimate of the observational interannual L_vdP/dT_s is marked in grey shading. Each model is represented by a lowercase letter. The ensemble model means for the 21 models and the five better-performing models are shown in solid circles and black cross, respectively. The least-squares linear regression lines and correlation coefficients between the x-axis and y-axis variables are shown.