Transcript Slide 1
Aerosol Effects on Rainfall Patterns Yi Ming Geophysical Fluid Dynamics Laboratory Princeton, New Jersey Most anthropogenic aerosols are in the NH mid-latitudes and tropics MODIS annual-mean aerosol optical depth (AOD) Will the response be confined locally? Tropics •Small f •Upper-level wind constrained by angular momentum conservation •Cannot sustain large meridional temp. gradient (MTG) Mid-latitudes •Large f •Large MTG balanced by thermal wind Forcing EQ 30N 60N 90N 30N 60N 90N 30N 60N 90N Response EQ Forcing EQ Response EQ or 30N 60N 90N A rough sketch of a paradigm Forcing EQ 30N 60N 90N Stationary Rossby wave Response Surface albedo feedback Eddies EQ 30N 60N 90N •Local vs. non-local effects (validity? Ways to distinguish and quantify them?) •Zonal-mean vs. zonal asymmetry Design of mixed-layer experiments 292 +2.76 K GAS Surface Temperature (K) 291 290 +0.55 K 289 BOTH Control 288 -0.62 K (direct effect only) 287 286 -1.90 K 0 10 20 30 40 50 60 Model Year Ming and Ramaswamy (2009) 70 AERO Aerosol direct + indirect effects -2.1 W m-2 Sign of nonlinearity: BOTH (0.55 K) < AERO + GAS (0.86 K) Zonal-mean changes Zonal-mean responses to aerosols and greenhouse gases •Dipole pattern of tropical rainfall change; •Role of the thermodynamic control (C-C). Surface temperature (K) Precipitation (mm day-1) BOTH GAS SUM SUM BOTH AERO AERO GAS Zonal-mean (Hadley) circulation change Meridional stream function (109 kg s-1) (clockwise circulation is positive) Stronger ascent Weaker ascent How does the Tropical heat engine response to aerosol forcing, and why? Difference in atmospheric heat transport (PW) From the viewpoint of atmospheric energy transport, the response gives rise to a cross-equatorial heat flux from SH to NH. Radiative 0.5 cooling AERO 0.4 BOTH 0.3 Dumping Picking up Even the energy energy 0.2 North Pole kicks in. 0.1 0 -0.1 GAS -0.2 -90 -60 -30 0 30 60 90 Zonal-mean change in atmospheric energy transport (PW) GAS Dry Static AERO Total Total Dry Static Latent Est. latent Assumptions: •No change in flow •Small Latent Zonally asymmetric changes (tropics) Aerosol-induced changes in tropical circulation The thermodynamic argument (TA) (Held and Soden, 2006) For the entire globe (or the Tropics in isolation), Precip. P qs M c Convective mass flux Mixing ratio of water vapor in BL M c P 0.07Ts Mc P Clausius-Clapeyron (CC) scaling If one applies TA to greenhouse gases (GHG)induced warming, M c / M c P / P 0.07 ΔT ΔT s s -5 ~ -6 % K-1 1 ~ 2 % K-1 Allen & Ingram (2002) Held and Soden (2006) Stephens & Ellis (2008) Does this apply to aerosol cooling? The mixed-layer GCM simulations suggest M c / M c P / P 0.07 ΔT ΔT s s Stronger Tropical -3.0 % K-1 3.8 % K-1 mean circulation Number-crunching time! Percentage differences in variance of Mc (%) NH NH SH SH ZonalZonally- ZonalZonallymean assy. mean assy. Aerosols -16 9.3 30 16 Gases -6 -18 -7 -30 Both -24 -22 38 -9 •TA works well for GAS, but only partially for AERO. •The zonal-mean response is dominated by AERO. A drying trend over central-northern India during the second half of the 20th century JJAS rainfall (mm day-1 50 years-1) CRU IMR PREC/L UDEL South Asia – A region under global and regional changes AR5 historical emissions (Tg/yr) of SO2 Larmarque et al. (2010) How aerosols and greenhouse gases may affect the South Asian summer monsoon? Aerosols •Atmospheric heating enhances pre-monsoon rainfall (Lau and Kim, 2006); •Surface cooling and reduced Indian Ocean SST gradient weaken monsoon (Ramanathan et al., 2005; Chung and Ramanathan, 2006). Greenhouse gases •Slower tropical (especially Walker) circulation (Vecchi et al., 2006); •Nonetheless, increased rainfall due to higher moisture content (Ueda et al., 2006). Model physics and chemistry in the GFDL CM3 Model (used for AR5) •Aerosol-Liquid Cloud Interactions A prognostic scheme of cloud droplet number concentration (Ming et al., 2007) with an explicit treatment of aerosol activation at cloud base (Ming et al., 2006). •Convection Parameterization Move from the relaxed Arakawa-Schubert (RAS) in CM2 to the Donner deep convection scheme (Donner, 1993) and the University of Washington (UW) shallow convection scheme (Bretherton et al., 2003). By providing in-plume updraft velocity, the latter two are ideal for implementing aerosol/cloud microphysics. •Online aerosol transport •Tropospheric and stratospheric chemistry Attribution of the recent trend of the South Asian summer monsoon using CM3 historical simulations Linear trends of average JJAS rainfall over central-northern Indian (mm day-1) GG AERO All forcing CRU Are the simulated trends statistically significant? JJAS rainfall (mm day-1 50 years-1) Ensemble-mean Student’s t-test Natural variation Ensemble member Spatial pattern of linear trends of JJAS rainfall (mm day-1 50 years-1) CRU All forcing GG AERO Spatial pattern of linear trends of vertical velocity (hPa day-1 50 years-1) All forcing GG AERO Ascent defined as negative How Hadley and Walker circulations respond to green-house gases and aerosols? Climatology AERO GG All forcing Zonally asymmetric changes (boreal winter mid-latitudes) Change in surface temp. (K) Different spatial structures over N. Pac. & N. Atl. RFP (W m-2) Not obvious from “forcing” Change in 300-hPa u (m s-1) Δ(Ts) (K) Consistent with Ts Change in circulation SLP (hPa) 300-hPa eddy stream function (ESF) (106 m2 s-1) 500-hPa Z (10 m) El Nino-like? 500-hPa Z (10 m) 7 Warm Trop. Pac. minus 7 cold Trop. Pac. ΔTs (K) Lau (1996) Change in precip. (mm day-1) EX TR-W TR-E Setup of idealized model experiments •Dry hydrostatic spectral dynamical core; •T42 and 20 sigma-layers; •No topography; •Zonal-mean winds and temp. are nudged towards GCM simulations; •Forced with GCM-simulated diabatic heating. Change in 300-hPa ESF (106 m2 s-1) Idealized model (TR-W) Idealized model (TR-E) Idealized model (TR) Trop. heating NW-SE tilt Extratropical heating NE-SW tilt Conclusions Zonal-mean changes •In the deep tropics, drier NH & wetter SH •In the subtropics & extratropics, the rich-get-poorer •With the purpose of re-establishing interhemispheric energy balance; Zonally asymmetric changes •In the tropics, Hadley circulation sensitive to aerosols, Walker circulation controlled by the tropicalmean ΔTs; •In the wintertime extratropics, the importance of the stationary Rossby wave excited by the tropical rainfall change. Zonal-mean change in θ (K) Lower troposphere static stability AERO Δθz (K) (10-35°N) -0.6 GAS BOTH AERO+GA S 2.1 1.1 1.5 Zonal-mean change in u (m s-1) Lower troposphere vertical wind shear AERO Δuz (K) (10-35°N) 1.2 GAS BOTH AERO+GA S -0.3 1.5 0.9 Zonal-mean change in tropopause height (hPa) 2 K km-1 Height BOTH Instability – source of nonlinearity? Phillips criterion Latitude AER O+GA S