Climate effects of Southern African biomass burning aerosol

Transcription

1 Climate effects of Southern African biomass burning aerosol Robert Wood, University of Washington with Naoko Sakaeda, Philip Rasch, Duli Chand, Tad Anderson, Bob Charlson

2 Direct effect of aerosol layer on TOA SW radiagon Single sca1ering albedo (approx) Coakley and Chylek (1974) DRE < 0 (cooling) DRE > 0 (warming) Surface albedo

3 Elevated aerosol layers Photo courtesy Steve Abel, SAFARI Field Experiment 2000

4 Aerosol layers over clouds seen with CALIPSO over SE AtlanGc Ocean (13 Aug 2006) Chand et al., Nature Geosciences, 2009

5 Example from SE Pacific during VOCALS Free troposphere Elevated aerosol layer Stratocumulus cloud

6 Biomass burning fires (2006, monthly)

7 Mean values of AOD and direct radiagve effect of elevated aerosol layers (July Oct 2006/2007) AOD DRE atm DRE toa and low cloud cover RFE toa Chand et al., Nature Geosciences, 2009

8 For a given absorpgon (SSA), the radiagve forcing efficiency is determined primarily by cloud cover Chand et al., Nature Geosciences, 2009

9 ...but depends strongly upon single sca`ering albedo CriOcal cloud fracoon increases strongly with SSA (brighter albedo required for posiove DRE)

10 AbsorpGon: Single sca`ering albedo Frequency of occurrence Data from land areas during SAFARI 2000 field campaign single sca1ering albedo at 550 nm From Leahy et al. (2007), Geophys. Res. Le,., 34, L12814, doi: /2007gl029697

11 Direct radiagve forcing (DRF) AEROCOM Models (Schulz et al. 2006)

12 Inter model standard deviagon of aerosol direct radiagve forcing (AEROCOM, Schulz et al. 2006)

13 Aerosol all sky direct radiagve forcing ΔF all =ΔF clr + C(ΔF cld ΔF clr ) 5 20 o S, 10 o E 10 o W Inter model cloud cover variaoons explain 70% of the variance in the aerosol direct radiaove forcing (DRF) Model predicoon of cloud cover important for accurate quanoficaoon of aerosol DRF ΔF all Using data from AEROCOM study, Shultz et al. (2006)

14 Microphysical effects on clouds 83% of layers elevated above clouds 17% touching clouds Aerosol layers touching clouds Aerosol layers elevated above clouds ConstanOno and Bréon, Geophys. Res. Le,., 2010

15 Semi direct aerosol effects SimulaOons using CAM 3.0 with slab ocean Present day (PD) and no carbonaceous aerosol (NC) simulaoons No aerosol indirect effects cloud changes are responding to direct aerosol driven changes in meteorology Approach allows separaoon of RF changes due to direct effects, cloud cover and liquid water path (LWP) Results presented for July October when biomass burning aerosol loading from Southern Africa are significant

16 Carbonaceous aerosol properges Aerosol opocal depth change (PD NC) Aerosol loading and cloud fracoon Cloud fracoon Sakaeda et al., J. Geophys. Res., accepted

17 Aerosol SW absorpgon induces... Over ocean o S, 0 10 o E warmer FT above clouds 20 30% weaker subsidence Sakaeda et al., J. Geophys. Res., accepted

18 Changes in low cloud properges due to carbonaceous aerosols (July Oct) Sakaeda et al., J. Geophys. Res., accepted

19 Surface air temperature change (PD NC) Surface SW radiaove effect (mostly direct) [ o C] [W m 2 ] Sakaeda et al., J. Geophys. Res., accepted

20 PrecipitaGon and aerosol induced changes July October mean ReducGons in tropical precipitagon

21 Summary Biomass burning over southern Africa during July Oct results in large semi permanent elevated aerosol layers which advect over the SE AtlanOc stratocumulus sheet RepresentaOon of the radiaove effects of these layers is a major uncertainty in global models and depends upon geing clouds as well as aerosols correct. Aerosols appear to induce significant changes in the large scale environment (subsidence, temperature, SST, clouds)

22 Outstanding quesgons What is the climate response to aerosol forcing by biomass burning aerosols over the South AtlanOc? How does the ocean respond to strong surface forcing? How do clouds respond to meteorological changes driven by elevated aerosol layers How do clouds respond microphysically to biomass burning aerosols? Do we know enough about the aerosol absorpkon properkes when the aerosols advect over the SE AtlanKc? What might we do collecovely to address these quesoons?

23 A strawman field program Designed to observe key aspects of clouds and elevated biomass burning aerosols Also provides key measurements of stratocumulus to cumulus transioon in clouds over increasing SST

24

25 Community Atmospheric Model (CAM) SimulaGons (Naoko Sakaeda, Phil Rasch) Preliminary analysis of 20 year CAM simulaoons Present day AOD tuned to match CALIPSO measurements Figure shows change in low cloud cover (biomass burning aerosols no biomass burning aerosols)

26 biomass burning aerosol above cloud stratocumulus clouds MODIS Aqua RGB (enhanced) 13 Aug 2006 SE AtlanGc 500 km

27 AEROCOM Models (Schulz et al. 2006) Direct radiagve forcing for cloudy skies

28 Measurements of Lidar rago (S) From Anderson et al. (2000), J. Geophys. Res.

29 Retrieval: color rago method (CR) (Chand et al. 2006, J. Geophys. Res.) CALIPSO data, integrated a1enuated backsca1er at 532 and 1064 nm (γ 532 and γ 1064 ) Determine color raoo χ water = γ 1064 /γ 532 from layers classified as cloud (z < 3 km) Unobstructed liquid clouds should have χ = 1, and so deviaoons from this represent aerosols above clouds Use Beer Lambert law to obtain AOD of aerosol layer: =1 ideally, but use unobstructed cloud to calibrate Angstrom exponent

30 DepolarizaGon rago method (DR) (Hu et al. 2007, Chand et al. 2007) Use depolarizaoon δ of cloud layer, combined with its integrated a1enuated backsca1er γ, to derive AOD of overlying layer ExOncOon to backsca1er raoo for water clouds (19 sr) Self calibraoon coefficient

31 Cloud Aerosol Lidar and Infrared Pathfinder Satellite ObservaGons

32 CALIPSO lidar (CALIOP) Backsca1er profiles at 532 and 1064 nm Parallel and perpendicular polarizaoon for 532 nm channel

33 Aerosol opgcal thickness retrieval methodologies

34 Comparison of DR and CR aerosol opgcal depths assuming å = 2 for CR method Increasing Angstrom exponent DayOme results are similar assuming å = 2

35 Cloud layer top heights

36 Angstrom exponent for layers above cloud

37 Aerosol opgcal depth for layers above cloud (by month 2006) June July August Sep Oct Nov

38 AOD and winds at 600 hpa

39 Determining the direct radiagve effect of elevated aerosol layers above the partly cloudy boundary layer (Chand et al. 2009, Nature Geosciences)

40 RadiaGve transfer model DISORT radiaove transfer model Aerosol properoes needed are AOD (from CALIPSO), single sca`ering albedo (ω=0.85, Leahy et al. 2006), Angstrom exponent (CALIPSO), asymmetry factor (g = 0.62) Cloud properoes are cloud opgcal depth and cloud effecgve radius (MODIS), and cloud fracgon Ocean surface albedo = 0.06 Determine aerosol radiagve effect for clear sky, cloudy sky, and all sky (Jul Oct 2006/2007)

41 AbsorpGon of solar radiagon by aerosols σ a (10 7 m 1 ) σ s (10 6 m 1 ) ϖ Single sca1ering albedo Aitken mode parocle conc. Sca1ering coefficient AbsorpOon coefficient Single sca1ering albedo From Clarke and Charlson, 1985, Science

42 Wavelength dependence of ϖ Leahy et al. (2007) Aerosol is relaovely more absorpove at longer wavelengths From Bergstrom et al. (2007), Atmos. Chem. Phys.

43 Effect of aerosol upon radiagve fluxes AOD AbsorpOon DRE (TOA) RadiaOve forcing efficiency