Contents Aerosol–Cloud Interactions and Their Radiative Forcing Aerosol Physics and Chemistry Climatology of Stratospheric Aerosols Climatology of Tropospheric Aerosols Dust Observations and Measurements Role in Radiative Transfer Role in Climate Change Soot
Aerosol–Cloud Interactions and Their Radiative Forcing U Lohmann, ETH Zurich, Zürich, Switzerland Ó 2015 Elsevier Ltd. All rights reserved. This article is a revision of the previous edition article by S M Kreidenwis, volume 1, pp 40–47, Ó 2003, Elsevier Ltd.
Synopsis Aerosols are essential for cloud formation. Every cloud droplet needs an aerosol particle, called cloud condensation nucleus, for activation. Likewise ice crystals either form on a subset of aerosol particles that act as ice nuclei or form by homogeneous freezing of supercooled solution drops. Anthropogenic aerosols cause an indirect radiative forcing by modifying cloud properties. This indirect radiative forcing has uncertainties that are larger than for any other forcing.
Aerosols are essential for cloud formation. Every cloud droplet needs an aerosol particle, called cloud condensation nucleus (CCN), for activation. Likewise ice crystals either form on a subset of aerosol particles that act as ice nuclei (IN) or form by homogeneous freezing of supercooled solution drops (liquid aerosols that took up water). In the absence of aerosols, several 100% relative humidity (RH) would be necessary for cloud droplets to form by homogeneous nucleation from supersaturated water vapor. Due to the ubiquitous presence of CCN, the RH in water clouds hardly exceeds 101%. The situation is different for ice clouds because IN are sparse, and formation of cirrus clouds is dominated by homogeneous freezing of solution droplets. This takes place at relative humidities below 100% with respect to water but well above 100% with respect to ice (at 60 C the RH for homogeneous freezing is 150%). At the same time, removal of aerosols by clouds and precipitation is the largest sinks for aerosols with diameters <1 mm. Thus the lifetime of aerosols is strongly linked to that of clouds.
An aerosol is deﬁned as a disperse system with air as the carrier gas and a solid or liquid disperse phase or a mixture of both. In atmospheric science, it is common to use the term ‘aerosol’ just for the solid or liquid particles and neglect the carrier gas. Aerosol particles range from 1 nm to several hundred micrometers in diameter. Aerosol particles can be as large as cloud droplets or ice crystals. Whereas cloud droplets or ice crystals only occur in isolated patches, aerosol particles, especially the smaller ones, are rather homogeneously mixed in the atmosphere. Aerosols have a variety of different formation mechanisms. They are divided into primary and secondary particles. Primary particles are already emitted as aerosol particles (either liquid or solid) into the atmosphere. Primary aerosols can result from bulk-to-particle conversion, such as wind blown dust from arid regions, or emissions of pollen and spores by plants or can originate from combustion processes. Mineral dust is mainly emitted in arid regions. Its emission rate depends on wind speed, soil moisture and the bare soil fraction. Sometimes
Encyclopedia of Atmospheric Sciences 2nd Edition, Volume 1
Aerosols j Aerosol–Cloud Interactions and Their Radiative Forcing
liquid-to-particle conversion is used for the formation of seasalt aerosols. Sea-salt aerosols originate from bursting of air bubbles that reach the ocean surface. The bursting bubbles leave water droplets containing sea salt behind. Upon evaporation of the water droplet, a sea-salt particle is released to the atmosphere. Primary aerosols are mainly large particles of natural origin and thus account for the largest aerosol mass in the atmosphere. Primary particles are distinguished from secondary aerosols, which form within the atmosphere from precursor substances (gases) by gas-to-particle conversion and form small aerosol particles. Examples of secondary aerosols in the atmosphere are sulfates, nitrates and secondary organic aerosols. Secondary particles have both natural and anthropogenic sources. Gas-to-particle conversion requires a nucleation process because a new phase (liquid or solid) is formed from a supersaturated gas phase. The most common form of particle nucleation is binary nucleation where the aerosol forms from two gas phase precursors. Homogeneous selfnucleation of a single species does not take place in the atmosphere as that would require that the vapor pressure of a single species is supersaturated. For binary nucleation to take place, each species can be subsaturated if the mixture is supersaturated. A prominent example of binary nucleation is the reaction of n moles water vapor H2O (g) with m moles of gaseous sulfuric acid H2SO4 (g), resulting in the nucleation of liquid sulfuric acid aerosol particles: (H2O)n(H2SO4)m (aq), where aq denotes the aqueous phase. A schematic representation of aerosol binary nucleation of H2O and H2SO4 with subsequent growth to larger sizes is shown in Figure 1. The ﬁrst step is that a cluster of H2O and H2SO4 is formed in the gas phase. Once the cluster exceeds its critical size, nucleation occurs and a stable sulfuric acid aerosol particle is formed. The critical size of a cluster is the size at which the droplet can exist in equilibrium with its vapor phase. Once the particle is formed, also other species, such as organics of low volatility can condense onto the particle and participate in its growth. Low volatility is
necessary as otherwise the organics would not remain in the condensed phase and evaporate again. The aerosol particle also grows by coagulation with other particles. The newly formed aerosol particle needs to increase 1–2 orders of magnitude in size before it can act as a CCN and be involved in cloud formation.
Hygroscopic Growth Condensation of water vapor on aerosol particles (water uptake) is important for aerosol particles with an afﬁnity for water vapor, i.e., for hydrophilic or hygroscopic aerosols, such as sulfate, nitrate, sea-salt, or mineral dust particles. This swelling of aerosol particles needs to take place before aerosol particles can act as CCN. Water uptake can cause a phase change if the soluble aerosol was solid before. If the aerosol is already liquid, then water uptake just leads to further growth and dilution of the salt solution. The phase change of the solid soluble aerosol to a liquid aerosol is called deliquescence and does not require a nucleation process. The opposite process, the formation of a solid aerosol, is called efﬂorescence or crystallization and requires a nucleation process. Nucleation requires overcoming an energy barrier, which can only be achieved if prior to nucleation the nucleating substance is in a supersaturated state as in the case of binary nucleation of H2O and H2SO4. The growth of the aerosol is described in terms of the growth factor, which is the ratio of the actual aerosol diameter to its diameter at 0% RH. It is only deﬁned for RHs below 100%. Some aerosols such as sulfuric acid (H2SO4) remain liquid at all relative humidities (Figure 2) at atmospheric temperatures and pressures and just change their size according to RH. Other aerosols, predominantly salts such as ammonium sulfate or ammonium bisulfate, change their phase. At which RH this phase change occurs depends on the direction of the change in RH. In case of ammonium sulfate, the sudden increase in diameter by 50% occurs at 80% RH. This increase in
Figure 1 Schematic representation of the nucleation and subsequent growth process for atmospheric binary homogeneous nucleation of H2SO4 and H2O. Once stable clusters are formed, also other substances such as organics can take part in the growth process. Particles may grow to sizes large enough to act as CCN on which cloud droplets may form eventually. Figure reproduced with courtesy from Curtius, J., 2006. Nucleation of atmospheric aerosol particles. C. R. Phys. 7, 1027–1045.
Aerosols j Aerosol–Cloud Interactions and Their Radiative Forcing
Figure 2 Water uptake of sulfuric acid (H2SO4), ammonium bisulfate (NH4HSO4), and ammonium sulfate ((NH4)2SO4) aerosols expressed in terms of their growth factor, which is the ratio of the actual diameter Dp to the dry diameter at 0% RH (Dp0), as a function of RH. Figure reproduced with courtesy from Seinfeld, J.H., Pandis, S.N., 1998. Atmospheric Chemistry and Physics: From Air Pollution to Climate Change. Wiley, 1326 pp.
diameter goes hand in hand with a phase change from solid to liquid and is referred to as the deliquescence relative humidity. Here the salt dissolves in water. If RH is decreased, ammonium sulfate remains in the liquid phase until 40% RH, where it suddenly solidiﬁes. This RH is called the crystallization relative humidity. The growth factor of ammonium sulfate exhibits a hysteresis, meaning the value of the growth factor is not unambiguously determined by RH, but depends also on the history of the particle. Between 40 and 80% RH the liquid phase of ammonium sulfate is metastable. Energetically, the aerosol would prefer to be in the solid state but it is prevented from doing so because it ﬁrst needs to surpass the energy barrier for nucleation.
Cloud Droplet Formation Cloud droplet formation is not a nucleation process because cloud droplets form on soluble or hydrophilic aerosol particles, which already have taken up large amounts of water below 100% RH. Cloud formation is best described by the Koehler equation: e ðrÞ=es ðNÞ ¼ S ¼ 1 þ a=r b=r 3 
where s is the surface tension between water and air. At 273.2 K, s is 0.0756 N m1 and rw, the water density, is 1000 kg m3. The terms of the Kelvin equation other than the radius of the droplet are summarized in term a of the Koehler equation: a ¼ 2s/(r rwRvT). Kelvin’s equation describes that the equilibrium vapor pressure is larger over a droplet with radius r than over a plane or bulk surface. It inversely relates the critical radius for droplet formation to the necessary supersaturation. At S ¼ 1.01, a typical value found in the atmosphere, the critical radius of the droplet needs to be 0.12 mm. This can never occur by chance as it involves 0.25 million water vapor molecules. A more reasonable number of 20 water vapor molecules forming a cluster with a critical radius of 0.5 nm requires a saturation ratio S of 10, i.e., an RH of 1000%, which does not exist in Earth’s atmosphere. The second contributor to the Koehler equation, Raoult’s law, is given for a plane surface of water as follows: e ðNÞ=es ðNÞ ¼ n0 =ðn þ n0 Þ
where e* is the equilibrium vapor pressure over a solution consisting of n0 water molecules and n solute molecules. This equation shows that if the vapor pressure of the solute is less than that of water and if the total number of molecules remains constant, the vapor pressure over the solution is reduced in proportion to the amount of solute present. The vapor pressure reduction arises from solute molecules at the surface that limit the exchange of water molecules between the water surface and the overlying vapor to those places where water molecules occupy the surface. For dilute solutions and applied to droplets of size r, Raoult’s law can be approximated as: e ðrÞ=es ðrÞ ¼ 1 b=r 3 ;
where b ¼ 3 i Ms mw =ð4p ms rw Þ 
The competition between Kelvin’s equation and Raoult’s law is summarized in the Koehler equation. Figure 3 displays the Koehler curves, i.e., the equilibrium vapor pressure as a function of the droplet radius, for droplets containing different amounts of salt. Note that every salt particle with a dry radius rd has its own individual Koehler curve. Because
The Koehler equation describes the ratio between the equilibrium vapor pressure over a solution droplet e*(r) with radius r to the saturation vapor pressure over a plane surface of water es(N) with a and b as described below. es(N) depends exponentially on temperature (T) as described by the Clausius– Clayperon equation: des =dT ¼ Lv es = Rv T 2  where Rv is the speciﬁc gas constant of water vapor (Rv ¼ 461.5 J K1 kg1) and Lv is the latent heat of vaporization. The Koehler equation has two contributions, the increase in vapor pressure that is associated with the formation of the surface (Kelvin equation) and the decrease in vapor pressure due to soluble substances in water (Raoult’s law). The Kelvin equation is given as: es ðrÞ=es ðNÞ ¼ S ¼ expð2s=ðr rw Rv TÞÞ
Figure 3 Koehler curves for NaCl (solid lines) and (NH4)2SO4 (dashed lines) for droplets originating from salt particles with different dry radii rd. In addition, the Kelvin curve (common for all particles) is shown in olive and Raoult’s law for (NH4)2SO4 as dot-dashed lines.
Aerosols j Aerosol–Cloud Interactions and Their Radiative Forcing
Raoult’s law applies to the droplet volume, it dominates over the Kelvin effect, that depends on droplet size, at small radii and lowers the equilibrium vapor pressure at small radii below S ¼ 1. The Kelvin effect is negligible for droplets larger than 1 mm where all curves approach S ¼ 1. The peak in the saturation ratio called critical saturation ratio (Sc) and the corresponding critical radius (rc) can be obtained by differentiating the Koehler equation with respect to r and setting the derivative to 0: rc ¼ ð3b=aÞ1=2 ;
1=2 Sc ¼ 1 þ 4a3 =27b
Sc of a speciﬁc Koehler curve is the minimum saturation ratio that is required for the corresponding solution droplet to grow to cloud droplet size. Theoretically, it could grow even larger, but the growth becomes increasingly slower the larger the droplet and more efﬁcient growth mechanisms such as growth by collision-coalescence with other droplets will take over (see Clouds and Fog: Cloud Microphysics). If the droplet has grown to r > rc, it is called an activated droplet. All droplets that have a critical saturation ratio Sc < S, where S is the supersaturation reached in the ambient air, can thus be activated. If S < Sc , a deliquesced salt particle can only grow to the radius at which the Koehler curve takes the value S. The Koehler curve represents equilibrium conditions and therefore has some limitations of its applicability. Large particles have large equilibrium radii and may have insufﬁcient time to grow to their equilibrium size in clouds that do not last long, such as convective clouds. Figure 3 shows that the higher S, the more and the smaller aerosols can be activated. It also shows that the largest aerosols are the best CCN because they require the smallest supersaturations to be activated. However, there are only few of them available. Also, because the diffusional growth depends inversely on the size of the droplets with the smallest droplets growing the fastest (see Clouds and Fog: Cloud Microphysics), the large aerosols may not reach their critical size by diffusional growth. On the other hand, the smallest aerosols require higher supersaturations to activate than exist in the atmosphere. Thus, these aerosols are not efﬁcient for cloud formation either. Mainly accumulation and coarse mode, and partly Aitken mode aerosols act as CCN and get activated into cloud droplets.
Ice Crystal Formation Once a cloud extends to altitudes where the temperature is below 0 C, ice crystals may form either by freezing of a cloud droplet or by direct deposition of the vapor to the solid phase. Both, freezing of cloud droplets and deposition of vapor to the solid phase, are nucleation processes because a stable cluster of the new ice phase has to form within the parent phase (vapor or liquid). Both homogeneous and heterogeneous nucleations are possible for the formation of ice crystals and have been observed in the atmosphere. Homogeneous freezing of liquid aerosols is the dominant freezing process in cirrus clouds. These liquid aerosols, also called supercooled solution droplets, are in equilibrium with the ambient RH below 100% with respect to water, but are not activated as cloud droplets. Heterogeneous ice nucleation, which requires the presence of
an IN, is the dominant ice nucleation process in mixed-phase clouds. Mixed-phase clouds are clouds that exist between 0 and 40 C and consist of a mixture of supercooled cloud droplets and ice crystals. IN are aerosol particles that provide a surface onto which water molecules can stick, bond together, and form aggregates with an ice-like structure. Differently from CCN, only a small fraction (one in 100 0001 000 000) of all aerosol particles can serve as IN at temperatures warmer than 40 C. Therefore, the criteria for aerosols to act as IN are less well understood as compared to the CCN ability of aerosol particles. Homogeneous deposition nucleation of vapor to form ice crystals is analogous to homogeneous cloud droplet nucleation. Also here the vapor pressure increase over an ice sphere is increased as compared to bulk ice, the more the smaller the crystal is. It can thus be described with Kelvin’s equation (see above) except that the parameters related to liquid water need to be replaced with those of ice. Homogeneous deposition nucleation is even less likely than homogeneous cloud droplet formation because of the higher surface tension between ice and air than between water and air and because the ice density is lower than the water density. Both differences increase the vapor pressure over an ice sphere more than that over a cloud droplet (see eqn ). On the other hand, homogeneous freezing of ice within a liquid drop is observed in the atmosphere. It occurs when statistical ﬂuctuations of the molecule arrangement of water produce a stable, ice-like structure, called ice germ. The formation of the surface of an ice-like structure requires an energy barrier to be overcome analogous to the crystallization of salts discussed above. If the ice germ exceeds the energy barrier for nucleation, it can grow spontaneously and cause the entire droplet to freeze. Experimental data on the freezing of pure water show that cloud droplets below a radius of 5 mm will freeze spontaneously at temperatures below 38 C Pruppacher and Klett (1997). Larger droplets freeze at slightly warmer temperatures. Homogeneous freezing of pure water only occurs at or above water saturation (Figure 4) since only at these conditions pure water droplets can exist in equilibrium. In cirrus clouds, the RH is not high enough to activate cloud droplets. The RH is, however, high enough for unactivated solution droplets to exist, which also can freeze homogeneously. Homogeneous freezing of solution droplets is the dominant pathway of ice crystal formation in cirrus clouds. Solution droplets refer to sulfuric acid or other liquid aerosols that have taken up water. The presence of salt causes the freezing point to be colder than that for pure water (freezing point depression). If the RH is not sufﬁciently high, the salt is too concentrated inside the solution droplets for freezing to occur. Therefore, homogeneous freezing of solution droplets occurs only at or above the dashed line that starts at 35 C in Figure 4. Homogeneous freezing of solution droplets deviates from the water saturation more and more with decreasing temperature below 35 C because solution droplets can freeze with higher salt concentrations and concentrated droplets can exist in equilibrium at lower relative humidities. Heterogeneous freezing initiates freezing with the help of an IN. An IN favors freezing over homogeneous freezing because it reduces the energy barrier to the formation of a critical ice germ.
Aerosols j Aerosol–Cloud Interactions and Their Radiative Forcing
Figure 4 Schematic of the main freezing processes as a function of temperature and the saturation ratio with respect to ice (Hoose, C.,Möhler, O., 2012. Heterogeneous ice nucleation on atmospheric aerosols: a review of results from laboratory experiments. Atmospheric Chemistry and Physics 12, 9817–9854). The solid line refers to saturation with respect to water. The dashed line that starts at 35 C refers to the homogeneous freezing line of solution droplets according to Koop, T., Luo, B., Tsias, A., Peter, T., 2000. Water activity as the determinant for homogeneous ice nucleation in aqueous solutions. Nature 406, 611–614.
Four heterogeneous freezing modes are distinguished in the literature (Figure 4). Immersion freezing refers to the freezing that is initiated from within the droplet. It requires that the IN is already immersed in the cloud droplet at warmer temperatures. Upon cooling freezing is initiated (Figure 4). Sometimes condensation freezing is distinguished from immersion freezing. It is thought that condensation freezing refers to a different pathway in which the aerosol containing the IN starts from subsaturated conditions. When water saturation is exceeded, a liquid phase is formed, which is large enough for an ice germ to form inside of it. This ice germ within the liquid phase then initiates the freezing (Figure 4). Contact freezing refers to the collision of an IN with a supercooled cloud droplet. It requires the presence of a cloud droplet and is therefore shown on the water saturation line in Figure 4. Deposition nucleation refers to the direct deposition of vapor onto an IN. It requires that the air is supersaturated with respect to ice (Si > 1). Deposition nucleation is important for cirrus clouds, when vapor is deposited for instance onto mineral dust particles that act as IN. It does not seem to be important for mixedphase clouds, because observations reveal that the majority of clouds with 0 and 40 C have liquid water droplets at the cloud base, suggesting that immersion or contact freezing are the most important freezing mechanisms in mixed-phase clouds. The criteria for aerosols to act as IN are thought to be (1) solid state, (2) size, (3) lattice structure, (4) molecular bindings with water, and (5) active sites. There is, however, some dispute about the importance of each of them and how many criteria are required. In order to reduce the energy barrier of heterogeneous nucleation below that of homogeneous nucleation, an ice germ has to form on a solid surface. A solid surface is necessary and is readily fulﬁlled for the most commonly occurring IN such as mineral dust, biological particles, and
soot. But also crystalline ammonium sulfate and some organics that are solid at certain conditions of temperature and RH can act as IN. Similarly to CCN also for IN size matters. The larger the IN surface, the larger the probability that a critical cluster forms on it. Also the larger the IN surface, the larger the probability that a good active site (see below) can be found onto it. A crystalline structure, which is similar to the ice lattice, is preferable for an IN. This is the case for silver iodide, which has been found to nucleate ice at temperatures up to 6 C. However, organics and mineral dust particles can also act as IN, although their lattice structures differ from that of ice. Thus, it is conceivable that the ability to form hydrogen bonds with water is more important than the lattice structure. Active sites refer to imperfections on the surface of the IN. They can be thought of as crevasses or steps in the lattice structure. The critical ice germ needs a smaller mass to reach the critical ice germ radius in crevasses or in corners of steps than on a plane surface, i.e., the energy barrier is particularly small in imperfections. Once ice is formed, additional water molecules can readily be attracted. Active sites are also used as an explanation why at a given temperature, RH and size of a given aerosol compound, only a fraction of a given aerosol species acts as IN. In summary, some combinations of size, lattice structure, molecular binding, and low interfacial energy with ice as promoted by active sites accounts for the IN ability of a substance.
Radiative Forcing of Anthropogenic Aerosols due to Aerosol–Cloud Interactions Aerosol particles can affect the climate by scattering and absorption of radiation and thus exert a radiative forcing due to aerosol–radiation interactions (RFari). In addition, aerosol particles can cause a radiative forcing by acting as CCN and IN, i.e., due to aerosol–cloud interactions (RFaci). Both RFari and RFaci have partially offset the greenhouse gas warming since preindustrial times. Whereas radiative forcing means that the atmospheric state remains constant, the cloud lifetime, coverage, or phase may change when perturbed by aerosols from anthropogenic activity. Adjustments in clouds occur on much faster time scales than the warming due to greenhouse gases. Therefore, it has been proven useful to also introduce an effective radiative forcing where macroscopic adjustments (cloud height, lifetime, and cover) to microphysical perturbations are considered as well. RFaci refers to an increase in cloud droplet number concentration resulting from an increase in anthropogenic aerosols. If the cloud water cloud remains constant, then the surface area of the cloud is larger and therefore more solar radiation is reﬂected back to space. This effect was previously referred to as the Twomey effect. RFaci is reported as the global annual mean change in the net top-of-the-atmosphere radiation since preindustrial times due to the aerosol-induced changes in cloud optical properties. RFaci has been estimated mainly from global climate models to have caused an RFaci of 0.7 W m2 with a range between 0.3 and 1.8 W m2 (Forster et al., 2007). Evidence for RFaci can be seen in so-called ‘ship tracks’ that leave behind white lines in satellite pictures. They are caused by the increase in albedo due to the injection of pollution aerosols from the ships that increase CCN and thus the number of cloud droplets. However, ship tracks cannot be
Aerosols j Aerosol–Cloud Interactions and Their Radiative Forcing
seen all the time because of counteracting processes. This includes a faster evaporation of the more but smaller cloud droplets or increased entrainment of warm dry air from above the boundary layer into these ship tracks. Buffering processes are also the reason why ERFaci is even more uncertain than RFaci itself. The total anthropogenic aerosol effect due to aerosol–radiation and aerosol–cloud interactions (ERFaci-ari) has been estimated from global climate models and satellite studies and amounts to 1.2 W m2 with a range between 0.2 and 2.3 W m2 since preindustrial times (Denman et al., 2007). Not only are the adjustment processes less well known, but they are also not well represented or are even missing in current climate models (Stevens and Feingold, 2009). One source of uncertainty is the ice phase. Compared to warm clouds and CCN activation, aerosol effects on mixedphase and ice clouds are much less understood (Lohmann and Feichter, 2005). While there is a consensus that mineral dust particles are good IN because they initiate ice formation at rather warm temperatures/low supersaturations, studies disagree about the importance of carbonaceous aerosols to act as IN (Hoose and Möhler, 2012). Terrestrial biogenic aerosols such as bacteria, pollen, and fungal spores have been identiﬁed as being good IN (Hoose and Möhler, 2012), and there are indications that marine planktonic diatoms also act as IN. These biological IN initiate ice formation at higher temperatures than mineral dust, but it is not clear yet if sufﬁciently large concentrations can be found in the atmosphere to substantially inﬂuence mixed-phase and ice clouds. Consequently, the contribution of anthropogenic emissions on IN concentrations is not yet determined. If more IN exist due to anthropogenic activity, supercooled clouds would freeze more readily. Because the vapor pressure of ice is lower than that of water, the ice crystals would grow at the expense of the cloud droplets (see Clouds and Fog: Cloud Microphysics). Glaciated clouds precipitate more readily which decreases their lifetime. This could partly offset the aerosol effect on warm clouds. However, anthropogenic emissions may deactivate IN due to coating with soluble material, such as sulfuric acid. Whether the faster glaciation or the deactivation
of anthropogenic IN is more important is still an open question. It differs in different climate models and is therefore partly responsible for the large range in ERFaciþari (Lohmann et al., 2010). In summary, aerosol–cloud interactions are not well known yet and their radiative and adjusted forcings will remain uncertain for some time to come.
See also: Aerosols: Climatology of Stratospheric Aerosols; Climatology of Tropospheric Aerosols; Observations and Measurements; Role in Climate Change; Role in Radiative Transfer. Clouds and Fog: Cloud Microphysics.
References Curtius, J., 2006. Nucleation of atmospheric aerosol particles. Comptes Rendus Physique 7, 1027–1045. Denman, K.L., Brasseur, G., Chidthaisong, A., et al., 2007. Couplings between changes in the climate system and biogeochemistry. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY. Forster, P., Ramaswamy, V., Artaxo, P., et al., 2007. Radiative forcing of climate change. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.), Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, pp. 129–234. Hoose, C., Möhler, O., 2012. Heterogeneous ice nucleation on atmospheric aerosols: a review of results from laboratory experiments. Atmospheric Chemistry and Physics 12, 9817–9854. Koop, T., Luo, B., Tsias, A., Peter, T., 2000. Water activity as the determinant for homogeneous ice nucleation in aqueous solutions. Nature 406, 611–614. Lohmann, U., Feichter, J., 2005. Global indirect aerosol effects: a review. Atmospheric Chemistry and Physics 5, 715–737. Lohmann, U., Rotstayn, L., Storelvmo, T., et al., 2010. Total aerosol effect: radiative forcing or radiative ﬂux perturbation? Atmospheric Chemistry and Physics 10, 3235–3246. Pruppacher, H.R., Klett, J.D., 1997. Microphysics of Clouds and Precipitation. Kluwer Academic Publishers, Dordrecht, The Netherlands, 954 pp. Seinfeld, J.H., Pandis, S.N., 1998. Atmospheric Chemistry and Physics: From Air Pollution to Climate Change. Wiley, NJ, 1326 pp. Stevens, B., Feingold, G., 2009. Untangling aerosol effects on clouds and precipitation in a buffered system. Nature 461, 607–613.