Clouds, Haze, and Climate Change Jim Coakley College of Oceanic and Atmospheric Sciences
Earth s Energy Budget and Global Temperature Incident Sunlight 340 Wm -2 Reflected Sunlight 100 Wm -2 Emitted Terrestrial Radiation 240 Wm -2 Absorbed Sunlight 240 Wm -2 T e = 255 K = 18 C Absorbed Sunlight = Emitted Radiation
Radiative Forcing Absorbed Sunlight = Emitted Radiation β (Change in Absorbed Sunlight Change in Emitted Radiation) = Change in Global Mean Surface Temperature Change in Absorbed Sunlight Change in Emitted Radiation = Radiative Forcing β Radiative Forcing = Change in Global Mean Surface Temperature β = 0.3 C/Wm -2 for the no-feedback radiative equilibrium model. β = 0.6 C/Wm -2 feedback. for climate models with the water vapor
Visible and Infrared Images of the Earth 0.65 μm 11 μm GOES West for 10 Oct. 07 1500Z
Cloud Radiative Forcing Change in Absorbed Sunlight Change in Emitted Radiation = Radiative Forcing F F A C = ( 1 A ) C F Cloud-Free + A = Radiative Flux - Reflected Sunlight, Emitted Infrared (Wm = Fractional Cloud Cover C F Overcast -2 ) Cloud Radiative Forcing = F F = AC ( F F ) F and F Cloud-Free are observed! Cloud-Free Overcast Cloud-Free
Cloudy Skies, F Reflected Sunlight Cloud-free, F Cloud-Free Observations from CERES aboard Terra for March 2001. High thick clouds in the tropics reflect strongly compared with the surrounding, largley cloud-free subtropics. Low-level clouds at midlatitudes also reflect strongly.
Cloudy Skies, F Emitted Longwave Radiation Cloud-free, F Cloud-Free Observations from CERES aboard Terra for March 2001. High thick clouds in the tropics emit at very cold temperatures compared with the surrounding, largely cloud-free subtropical regions.
Cloud Radiative Forcing Albedo Cloudy 0.30 Sunlight Absorbed (Wm -2 ) Infrared Radiation Emitted (Wm -2 ) 240 240 Cloudfree Net Forcing 0.15 289 270 49 30 Radiative forcing = Change in absorbed sunlight minus change in emitted radiation. Cloud Forcing = 19 Wm -2 Clouds lower the Earth s Temperature by 19 Wm -2 0.6 C/(Wm -2 ) ~ 12 C.
Amounts of Water in Cloud Droplets and as Vapor within the Cloud Layer Typical number of droplets, 100 cm -3 Radius ~ 0.01 mm (10 μm) Volume = 4 3 9 πr 3 = 4 10 Density of water = 1 g cm -3 cm 3 Mass of droplets = 100 4 10-9 g cm -3 = 4 10-7 g cm -3 Mass of droplets for 100 m thick cloud = 0.004 g cm -2 Amount of water in the atmosphere 2 g cm -2 1-km thick cloud in the lower troposphere contains ~.04 g cm -2 in condensed water compared with ~1 g cm -2 in vapor. A 10% error in the amount of water available for condensation leads to greater a 100% error in liquid water large errors in cloud albedo.
1980 s Grid Sizes Used in Climate Models 1990 s Spatial resolution used within climate models is woefully inadequate to deal with the scales over which clouds change significantly (tens to hundreds of meters). Modelers attempt to mimic the statistical behavior of clouds through parametric models cloud parameterization schemes. 2000 s Source: IPCC (2007)
Anything to do with water vapor and clouds. Typical ~100 km model grid box (in 2006) Models must predict how much water condenses into droplets and freezes into ice crystals to predict the amount of sunlight reflected and infrared radiation absorbed and emitted.
Sensitivity of Climate Models Climate sensitivity in 19 climate models for cloudfree simulations, 0.5 C/Wm -2 (open circles). Climate sensitivity including cloudfreedbacks, range from 0.4 1.2 C/Wm -2. Source: Cess et al. (1990)
Observed and Predicted Global Surface Temperature Changes Colored lines indicate predicted temperatures associated with different paths of economic and technological development called scenarios. Gray shaded regions indicate range of uncertainty in temperature predictions due to range of predictions by different climate models and some observational constraints. Source: IPCC (2007)
Part II: Effects of Haze on Clouds
Radiative Forcing Since 1750 Source: IPCC (2007)
Aerosol Indirect Radiative Forcing (also referred to as the Twomey effect) For a given amount of condensed water Pristine clouds have fewer droplets and the droplets are larger. Clouds polluted by haze particles have more droplets and higher reflectivities even though the droplets are smaller.
Additional Aerosol Indirect Effects For constant liquid water amounts, smaller droplets leads to higher cloud reflectivies (Twomey 1972) Drizzle formation suppressed in clouds with smaller droplets (Albrecht 1989): Polluted clouds contain more liquid water than unpolluted clouds. Polluted clouds will have longer lifetimes. Polluted clouds will give rise to higher fractional cloud cover. Net amplification of aerosol indirect forcing in climate models estimated to be twice that predicted assuming constant liquid water amount.
Cloud Droplet Radius and Optical Depth and Aerosol Column Number Droplet Radius Optical Depth 4-km AVHRR data for 1990. Trends between cloud properties and aerosol burdens shown for month of daily average values composited for 17.5 17.5 latitude-longitude regions. Blues indicate negative and Reds positive correlations. Source: Sekiguchi et al. (2003)
Changes in Aerosols near Clouds 1-km Terra MODIS Data 0.55-μm Aerosol Optical Depth Sun Glint Boundary Fine Particle Fraction Cloud contamination of the cloud-free Fine regions used for aerosol retrievals. Growth in size with increase in relative humidity. Mixed New particle production. Growth in aerosol burdens due to cloud processing Coarse of aerosols. Enhanced illumination of the atmosphere due to the presence of the clouds. Clearly, the behavior of aerosols near clouds requires investigation.
Ship Tracks and the Indirect Effect of Aerosols Aqua 1-km MODIS 1915 UTC 11 June 2002 2.1-μm Affected clouds can be compared with nearby pristine clouds under conditions in which the only difference is that of the additional particle loading. Ship tracks revealed that polluted clouds have higher reflectivities at visible wavelengths than nearby unpolluted clouds (Coakley et al. 1987).
Analysis of Ship Tracks in MODIS Data 1-km Terra MODIS 22 July 2001, 1945 UTC 2.1 μm New retrievals account for partial cloud cover within 1-km MODIS fields of view (Coakley et al. 2005). Droplet radii derived using 1.6, 2.1, and 3.7-μm MODIS channels. New semi-automated track finding scheme allows comparisons between clouds separated by only a few kilometers.
Change in Optical Depth for Clouds with Large Change in Droplet Radius NOAA-14 June 1999 Source: Coakley and Walsh (2002)
Liquid Water Amounts for Polluted and Unpolluted Clouds Unlike climate model simulations, polluted clouds have less liquid water than nearby unpolluted clouds Source: Coakley and Walsh (2002)
Simulated Cloud Albedos Y (km) 6 5 4 3 2 1 0 Particles 40 cm -3 Particles 75 cm -3 Droplets 13 cm -3 Droplets 40 cm -3 0 1 2 3 4 5 6 X (km) 6 Particles 150 cm -3 Droplets 89 cm -3 Y (km) 6 5 4 3 2 1 0 0 1 2 3 4 5 6 X (km) 6 Particles 300 cm -3 Droplets 167 cm -3 Large eddy simulation of marine stratocumulus. 6.5 6.5 1.5 km 3 domain 50 m resolution droplet number and sizes calculated radiative heating incorporated 5 5 Y (km) 4 3 2 Y (km) 4 3 2 Partly cloudy 1-km field of view 1 1 0 0 1 2 3 4 5 6 X (km) 0 0 1 2 3 4 5 6 X (km) Source: Ackerman et al. (2003)
Ship Track Under Dry Troposphere Head Dry Moist
Summary of Findings from Ship Tracks Droplet growth is inhibited in polluted clouds (consistent with hypothesis that droplet growth and drizzle formation is suppressed in polluted clouds). For overcast conditions, polluted clouds lose liquid water (inconsistent with suppression of drizzle hypothesis but consistent with LES model results for sufficiently dry overlying tropospheric air). Marine stratus appear to dissipate through drizzle which may be supported by relatively moist overlying tropospheric air (consistent with LES model results). Under broken cloud conditions, haze pollution causes increases in cloud cover, cloud optical depths, and liquid water amounts (consistent with hypothesis that pollution suppresses drizzle formation and loss of liquid water). Entrainment rates appear to be suppressed for afternoon clouds (liquid water amounts for morning clouds appear to be more responsive to pollution than those for afternoon clouds). Estimates of the aerosol indirect radiative forcing will depend on how accurately changes in cloud fraction, droplet radius, and cloud liquid water can be predicted.
Ship Tracks off the Coast of France Aqua MODIS True Color Image, 25 March 2003