Chapter 1

Aerosols, Clouds and Chemstry-Climate Interaction

* The first is the demonstration by model studies that the increase in anthropogenic sulfate particles is potentially a large contributor to the decadal change in the global radiative forcing of climate (Figure 1). Increases in sulfate aerosols can lead to a radiative cooling directly by enhancing clear sky albedo and indirectly through modification of the shortwave cloud forcing.

The magnitude of the estimated direct and indirect global radiative cooling effects of sulfates during the last century (~ -1 Wm^-2), and more importantly, the uncertainty in the estimated global effects
(0 to -3 Wm^-2) are so large (Figure 1) that aerosols now rank with cloud feedback and ocean-atmosphere interactions as the major barriers toward a reliable prediction of climate changes on decadal and longer time scales.

Figure 1. Potential anthropogenic radiative forcing of climate. Direct effect refers to the increase in clear-sky albedo. The indirect effect is the increase in cloud albedo. (IPCC, 1994)

* A parallel development is the demonstration that the global radiative effects of other continental aerosols (e.g., mineral dust and carbonaceous particles) may be comparable to the sulfate effects [Andreae, 1995; Li et al., 1995]. These aerosols are expected to have a particularly large effect in reducing the solar radiation over the tropical and sub-tropical oceans, because of the proximity of these oceans to arid regions [Prospero, 1979 for mineral dust] and equatorial forests [Andreae, 1995 for black carbon]. The radiative effects of these background aerosols is an important factor for simulating the present day climate.

* Another major emerging issue (with implications for decadal to longer time scale changes in anthropogenic climate forcing) is the link between regional emissions of reactive gases and their global impacts on the radiatively active species such as tropospheric ozone [Intergovernmental Panel on Climate Change (IPCC), 1990]. In this regard, a fundamental issue concerns the inter-hemispheric mixing of continental air. The gaseous pollutants and continental aerosols from the northern hemisphere encounter the relatively pristine air from the southern hemisphere in the vicinity of the Intertropical Convergence Zone (ITCZ). The rapid vertical transport within the deep convective-cirrus cloud systems of the ITCZ and subsequent horizontal distribution by fast zonal upper air flows provide a fundamental link between regional emissions and global distributions of radiatively active species. Our current understanding of the role of the ITCZ in the transport of trace gases and aerosols is largely based on model studies.

* Several recent studies [Ohmura and Gilgen, 1993; Wild et al., 1995; Cess et al., 1995; Pilewskie and Valero, 1995; Ramanathan et al., 1995; and Gleckler and Weare, 1995] have discovered a major gap in our understanding of the magnitude of the solar radiation absorbed within the atmosphere and that reaching the surface.

The Ohmura and Gilgen (1993) study uses the GEBA (Global Energy Balance Archive, World
Climate Program) data set. The global annual mean shortwave solar radiation is estimated at
142 Wm^-2, which is significantly smaller than model estimates of net surface solar radiation. Figure 2 adopted from Wild et al. (1995) compares the 142 Wm^-2 estimated by GEBA with 9 general circulation models (GCMs) which fall in the range of 164 to 184 Wm^-2. The systematic discrepancy cannot be attributed to insufficient cloudiness, since these GCMs simulate the top-of-atmosphere net solar radiation (incoming minus reflected) within a few Wm^-2. The Gleckler and Weare study, which employs the Comprehensive Ocean - Atmosphere Data Set (COADS), focuses on just the world oceans and comes to a similar conclusion: "Preliminary comparison of zonal average fluxes suggest that most atmospheric general circulation models produce excessively large ocean surface fluxes of net solar heating."

What is the implication to atmospheric solar absorption?

By combining GEBA data with satellite radiation budget data, Ohmura and Gilgen (1993) estimate atmospheric absorption to be 97 Wm^-2, whereas GCMs (Figure 2) would yield values less than 75 Wm^-2. In conclusion, the model atmosphere absorbs significantly less radiation than the real atmosphere, which is compensated by increased solar absorption at the surface.

Is the discrepancy in the clear or in the cloudy atmosphere?

or

Is it due to observational uncertainties?

Figure 2. Absorbed solar radiation at the surface: model vs. pyrenometer observations.

For nearly four decades, a systematic discrepancy has existed between the model clouds and the observations [Stephens and Tsay, 1991; Liou, 1992]. Model clouds are brighter and less absorptive than observed clouds. As summarized by Liou (1992): "Reflectance and absorptance computed from theoretical programs . . . . are generally higher and lower, respectively than observed data. The largest cloud absorptance derived from theoretical calculations has an upper limit of ~ 20% . . . However, the observed cloud absorptance could be larger than 30%." Three recent observational studies [Cess, et al., 1995; Pilewskie and Valero, 1995; and Ramanathan, et al., 1995], employing independent instruments and independent measuring platforms (including high altitude aircraft, satellites, ships, buoys and atmospheric soundings), have examined this issue for cloud systems in the tropical Pacific. They conclude that cloudy skies in the tropical western Pacific absorb about 8% more (or about 35 Wm^-2 more) of the top-of-atmosphere solar insolation when compared with clear sky values. The models predict, on the other hand, that cloudy skies absorb the same as clear skies. The issue, however, is far from settled [Stephens, 1996; Cess and Zhang, 1996; Pilewskie and Valero, 1996]. In the mean time, it has been suggested [Byrne et al., 1996; O'Hirok and Gautier, 1996] that cloud morphology and three-dimensional structures not accounted for by plane parallel theory can enhance solar absorption.

Some of the difference may be due to missing processes in the clear atmosphere which include [Li et al., 1995; Stephens and Tsay, 1991] absorbing aerosols, super-lorentzian water vapor lines, enhanced water vapor continuum absorption.

Clearly there is mounting evidence that climate models are missing a large heat source in the atmosphere (at least in the tropics), and are compensating for it by overestimating the solar energy absorbed by the sea surface and the land surface.

In addition, we need to examine how continental aerosols (particularly dust, black carbon and organics) and soluble species modify the clear and cloudy atmospheric solar absorption in the tropics. One basic flaw in our models, as well as in our general approach to the cloud-radiation interaction problem, is that we assume radiative properties of pure water and ice without any consideration for the radiative effects of absorbing aerosols or chemical species within cloud drops and ice crystals.

We cannot also entirely rule out the possiblilty that some of the differences are due to intrumentational error or other uncertainties in the observations. Accurate and calibrated observations are required to settle the issue.

As described next, the equatorial Indian Ocean, during a brief period between January and April of each year, performs a unique natural experiment that enables us to address all of teh above issues simulataneously.