Jost Heintzenberg1and Robert J Charlson2

1Leibniz Institute for Tropospheric Research, Leipzig, Germany 2Department of Atmospheric Sciences, University of Washington, Seattle, WA, U.S.A.

Clouds populate the Earth's atmosphere from the surface, as fog, to the mesosphere, as noctilucent clouds (cf. Table 1.1). They form whenever air is cooled sufficiently for its relative humidity to exceed 100%. Such cooling occurs, for example, when air is lofted upward or when a volume of air loses energy by radiating longwave radiation. Clouds can form at temperatures greater than 0°C (so-called "warm clouds") or below 0°C, and they can exist for lengthy periods of time as supercooled water droplets or as ice, either frozen droplets or crystals grown from the vapor phase. Since cooling can occur at almost any altitude and in myriad meteorological circumstances, clouds take on a nearly indescribable range of physical appearances, ranging from massive cumulus that dominate the sky to wispy veils that may even be too thin to be seen with the naked eye.

Clouds, however, constitute the largest source of uncertainty in the climate system, and there are solid reasons why our knowledge of clouds and their related processes is very limited. To approach these issues, this Ernst Strungmann Forum was convened to assess the limits of current knowledge and to offer new approaches to the understanding of cloud-related issues in the Earth system.

Perturbations of Clouds and Related Aerosols

Humankind is perturbing the Earth's cloud system through its actions (e.g., emissions and surface changes). Contrails, which result from aircraft emissions, represent the most obvious (but not necessarily most relevant) and easily perceived evidence of regional perturbations. Other anthropogenic cloud perturbations in the form of ship tracks, found in persistent low marine clouds, are clearly visible from space. Table 1.2 lists the primary mechanisms of

Table 1.1 Range of typical cloud properties. LWC/IWC = liquid water content/ice water content; N, hydrometeors

= number of cloud particles per volume of air.

Table 1.1 Range of typical cloud properties. LWC/IWC = liquid water content/ice water content; N, hydrometeors

= number of cloud particles per volume of air.

Location

Height (km)

(°C)

(cm 3)

Surface

0

Fog

= 0

10-100

1-100

Lower troposphere

1-5

Cumulus

10 to = -35

100-1000

10-1000

Lower troposphere

1-3

Stratus

10 to = -35

100-500

10-1000

Troposphere

1-15

Cumulonimbus

0 to -60

1000-10,000

100-1000

Upper troposphere

7-15

Cirrus

-40 to -90

1-10

0.01-10

Stratosphere

15-25

Polar stratospheric clouds

< -80

0.001-0.01

1-10

Mesosphere

80-85

Noctilucent clouds

= -120

0.00001-0.0001

25-500

anthropogenic perturbations of clouds recognized today (for a detailed discussion, see Chapters 6, 15-17).

Anderson et al. (Chapter 6) devoted considerable time to the discussion of confounding meteorological influences, which makes it difficult to test hypotheses of anthropogenic effects on clouds. They offer strategies for separating aerosol and meteorological effects in view of the classical "null" hypothesis combined with specific atmospheric settings in which potential anthropogenic cloud changes should be sought.

The observed and hypothesized perturbations of clouds listed in Table 1.2 require a comparison to long-term trends in observed clouds over the past several decades—a period marked by rapidly rising temperatures and changes in the Earth's radiation budget. The radiation budget controls the formation of clouds and is also strongly influenced by their existence, as discussed in Chapter 2. Here, Norris and Slingo review multidecadal variations in various cloud and radiation parameters, documented in previous studies; they argue that no conclusive results are yet available. Problems include the lack of global and quantitative surface measurements, the shortness of the available satellite record, the inability to determine correctly cloud and aerosol properties from satellites, many different kinds of inhomogeneities in data, and insufficient precision to measure the small changes in cloudiness and radiation, which nevertheless can significantly impact the Earth's climate. Their recommendations to improve this situation include (a) processing the available historical measurements as a means of mitigating inhomogeneities, (b) providing better

Table 1.2 Observed and hypothesized anthropogenic perturbations of clouds. CCN = cloud condensation nuclei; IN = ice nuclei; LWC = liquid water content; PBL = planetary boundary layer.

Cloud type

Perturbation

Potential mechanism

Contrails

+ Albedo

Water vapor and anthropogenic CCN/IN1

Contrails

- Daily temperature range

Change in air traffic in connection with 9/112

Ship trails

+ Albedo

Anthropogenic water vapor, and CCN3

Continental stratocumulus

+ Albedo

Anthropogenic CCN4

Continental stratocumulus

+ Cloud-top temperature

Anthropogenic CCN5

Continental stratocumulus

- Precipitation

Anthropogenic CCN6

Global PBL stratocumulus

+ Albedo

Anthropogenic CCN7

Continental rain clouds

- Precipitation

Anthropogenic CCN8

Continental deep convection

+ Freezing level

Anthropogenic CCN9

Continental low clouds

+ Precipitation

Cloud seeding with CCN or IN10

Continental

± Cloudiness

Surface flux change attributable

low clouds

to vegetation change11

Marine PBL clouds

- "Effective radius"

Anthropogenic CCN12

PBL stratocumulus

- LWC

Anthropogenic soot13

Cloud formation

+ Atmospheric heating

Anthropogenic greenhouse gases14

Global cloud cover

+ Cloudiness

Cosmic radiation, ions, anthropogenic CCN15

Regional weather

± Synoptic weather systems

Anthropogenic energy release or redirection16

1Scorer 1955, Meerkötter et al. 1999; 2Travis et al. 2002; 3Twomey 1974, Coakley et al. 1987; 4Twomey 1974, Krüger and Graßl 2002; 5Devasthale et al. 2005; 6Albrecht 1989, Rosenfeld 1999, Rosenfeld 2000; 7Twomey 1974, Nakajima et al. 2003, Sekiguchi et al. 2003; 8Bell et al. 2008; 9Andreae et al. 2004; "Garstang et al. 2004; "Pitman et al. 1999, Ray et al. 2003; 12Twomey 1974, Albrecht 1989, Han et al. 1994; "Ackerman et al. 2000; "Douville et al. 2002, Wetherald and Manabe 2002; 15Marsh and Svensmark 2000; "Hoffman 2002

1Scorer 1955, Meerkötter et al. 1999; 2Travis et al. 2002; 3Twomey 1974, Coakley et al. 1987; 4Twomey 1974, Krüger and Graßl 2002; 5Devasthale et al. 2005; 6Albrecht 1989, Rosenfeld 1999, Rosenfeld 2000; 7Twomey 1974, Nakajima et al. 2003, Sekiguchi et al. 2003; 8Bell et al. 2008; 9Andreae et al. 2004; "Garstang et al. 2004; "Pitman et al. 1999, Ray et al. 2003; 12Twomey 1974, Albrecht 1989, Han et al. 1994; "Ackerman et al. 2000; "Douville et al. 2002, Wetherald and Manabe 2002; 15Marsh and Svensmark 2000; "Hoffman 2002

retrievals of cloud and aerosol properties, and (c) extending the record farther back in time. In addition, they advocate an observation system with sufficient stability and longevity to measure long-term variations in cloudiness and the radiation budget with improved precision and accuracy. Unfortunately, as they note, there is currently little prospect in enhancing the present system, which is, moreover, in danger of deterioration since there are no definite commitments to replace several critical instruments when the current satellite missions end.

Clouds consist of particles of condensed water that have grown from either a cloud condensation nucleus (CCN) or an ice nucleus (IN), which caused either a supercooled water droplet to freeze by means of several possible mechanisms or water vapor to deposit directly to form solid water ice. Because CCN and IN are found in the form of aerosol particles, and because almost all aerosol particles can become CCN and some of them are inherently IN, understanding how and why clouds form and what properties they have requires us first to understand the nature and amounts of aerosol particles. The atmospheric aerosol spans a range of four orders of magnitude in particle size and seven orders of magnitude in number concentration (cf. Figure 1.1); CCNs are a subpopulation of this aerosol. Figure 1.1 illustrates the size and concentration ranges that typically act as CCN. Again, complexity arises because of the myriad sorts of aerosol particles, deriving from a host of natural and anthropogenic aerosol sources, that produce the starting material for the formation of cloud particles.

In Chapter 3, Kinne (Part 1) and Poschl et al. (Part 2) discuss climatologies of cloud-related aerosols in terms of particle number, size, and hygroscopic properties. To date, the high temporal and spatial variability of concentration, size, and composition of atmospheric aerosols has been mapped, based largely on insuffi ciently evaluated datasets of model simulations or satellite retrievals. Their approach merges data from ground-based remote-sensing networks into multi-model, median background fields that yield global monthly maps of columnar particle properties. The vertical distribution of aerosol characteristics is derived from global modeling. Applying the argument that hygroscopic growth of atmospheric aerosol particles is relatively well-constrained, global

Diameter D (nm)

Figure 1.1 Typical near-surface nonurban continental number-size distribution of atmospheric particles (Birmili et al. 1999; Heintzenberg et al. 1998). The typical size distribution of cloud condensation nuclei (CCN) is illustrated by the drop-scavenged fraction according to counterflow virtual impactor (CVI) data from Mertes et al. (2005). The CVI-scavenging data are extrapolated from their upper limit at D = 900 nm to the value of one at 10,000 nm. The question mark indicates the lack of data for smaller particles.

Diameter D (nm)

Figure 1.1 Typical near-surface nonurban continental number-size distribution of atmospheric particles (Birmili et al. 1999; Heintzenberg et al. 1998). The typical size distribution of cloud condensation nuclei (CCN) is illustrated by the drop-scavenged fraction according to counterflow virtual impactor (CVI) data from Mertes et al. (2005). The CVI-scavenging data are extrapolated from their upper limit at D = 900 nm to the value of one at 10,000 nm. The question mark indicates the lack of data for smaller particles.

monthly maps for concentrations of CCN are presented. The uncertainty of these results is not known.

Cloud characteristics have vertical and geographical variations, which are important but poorly constrained by present experimental methods. Isaac and Schmidt (Chapter 4) describe the in-situ and remote-sensing instrumentation currently available, as well as potential problems in discerning cloud properties. They discuss the necessity to measure parameters on the scales of interest and to present those measurements in proper units. Recommendations for future action include improvements in the accuracy of cloud measurements, global cloud data sets, and better collaborations between those who make and those who use in-situ and remote-sensing measurements.

Variability and potential trends of cloud properties affect not only the global radiation budget but also the global hydrological cycle through precipitation (e.g., rain and snow). Precipitation is difficult to assess on large scales. Based on recent developments in passive and active remote sensing, Takayabu and Masunaga (Chapter 5) review current understanding of extreme rainfall, as well as the statistics of light rain and rain from shallow clouds. They find a "butterfly" geographical pattern of shallow rainfall across the equator over both the tropical Pacific and Atlantic oceans. It is not fully understood why this quasi-symmetric pattern appears, as the tropical convergence zones, which geographically constrain deep convective rainfall, are highly asymmetric around the equator. The nature of extreme precipitation varies, depending on the timescale of interest, and is discussed in terms of hourly and daily extremes. Satellite observations imply that the global distribution of extreme precipitation shows a systematic difference from the total rainfall map in terms of, for example, the contrast between land and ocean. Results suggest that the realistic reproduction in models of synoptic systems as well as proper representations of shallow convection and its interaction with the synoptic-scale systems are indispensable for adequate reproduction of extreme daily precipitation.

Anderson et al. (Chapter 6) confirm the findings of the authors of Chapter 5 and emphasize that the most uncertain aspects in current knowledge concern ice microphysics and ice nucleation. Particular difficulties exist because of the confounding effects of built-in correlations of aerosols, clouds, and the meteorological fields in which they are found. These discussions strongly confirm the necessity of understanding and quantifying aerosol and cloud effects as a prerequisite to a full explanation of the climatic records of the twentieth century. Of considerable interest to the entire Forum was the conclusion that observational evidence for large-scale impacts of aerosols on cloud albedo, cloud amount, and precipitation remain ambiguous. Despite this ambiguity, participants were convinced that the emerging trend of warming over the past few decades makes it imperative to look for and quantify coincident changes in clouds. They emphasize the serious need for long-term planning of satellites to monitor the Earth's radiation budget and propose suggestions for new technologies and new orbits (e.g., at the Lagrange point L1 in space).

0 0

Post a comment