Chapter 2

Rationale: Why the Indian Ocean?

Strong Source of Natural and Anthropogenic Emissions

The Indian subcontinent and surrounding nations are rich sources for many kinds of aerosols such as mineral dust of natural and anthropogenically influenced sources, nitrates from agricultural and traffic-related sources, sea salt, sulfate particles and organic aerosols due to:

* gas-to-particle conversion of secondary oxidation products of organics emitted by the biosphere (vegetation, ocean);

* emission from industrial sources and transport vehicles;

* biomass burning (heating, cooking, shifting agriculture, agricultural waste burning) including optically and microphysically active carbonaceous aerosols; and

* primary aerosols, i.e., soil dust and sea-salt.

Figure 3. SO2 and CO2 emissions as a percentage of the global total for 1980, and CO2 emissions projections for 2025. 1980 SO2 values from Spiro et al., 1992; CO2 values from The UN Statistical Yearbook, 1994-5. 2025 projections from EPA Policy Options for Stabilizing Global Climate, Appendix B, 1985. The global projection for 2025, using EPA "Rapidly Changing World" scenario, is in excellent agreement with the "Business as Usual" scenario of the Intergovernmental Panel on Climate Change (IPCC).

S&E Asia: Bangladesh, Brunei, Burma (Myanmar), India, Indonesia, Malaysia, Pakistan, Philippens, Singapore, Sri Lanka and Thailand CP Asia: Cambodia, China, Korea, Laos, Vietnam

It is quite likely that mixed aerosol types are also very common, e.g., due to oxidation of SO2 to H2SO4, and H2SO4 (and HNO3) uptake by soil dust or sea salt, reducing the potential for new particle formation.

The equatorial Indian Ocean region provides us with a gigantic laboratory to observe and study the chemical interactions between gases and particulate matter, particulate matter and clouds, as well as the microphysical and radiative properties of the particles. Furthermore, this region provides a unique opportunity to observe the anthropogenic sulfate effects in the context of other equally important continental aerosols such as dust and carbonaceous particles including black carbon.

The importance of other aerosols for the global forcing can be inferred from Table 1. Of equal importance is the fact that the subcontinent is projected to be sufficiently industrially advanced that it may become a large source for CO2 (Figure 3), sulfate and other particulate and trace gas emissions [Wolf and Hidy, 1995]. Thus the Indian Ocean is in contrast to the tropical Atlantic, where industrially emitted chemicals are minimal.

Favorable Surface Flow for Testing the Direct Aerosol Effect

From January to April the predominant circulation in this region consists of a low level flow from the northeast (NE), i.e., from the polluted land in the north to the ocean in the south (Figure 4). From north to south, the low level air parcels witness a transition from clear skies to marine stratocumulus to the deep-convective-cirrus cloud systems of the ITCZ (Figure 5). This NE monsoon, or alternately the Asian winter monsoon, should facilitate the formation and transport of new sulfate particles, along with other pollutants (e.g., NOx and hydrocarbons) to regions of the ocean far away from urban centers.

Figure 4. Seasonal surface wind flow in the Indian Ocean. The thickness of the arrow is proportional to the force of the wind. From the Indian Ocean Atlas, 1976.

Figure 5. Asian Winter Monsoon. A schematic of the low level cross equatorial flow. (Courtesy of T.N. Krishnamurti)

Substantial quantities of dust and smoke particles from biomass burning will be transported as well. In addition, the period from January to April is the dry season in India and its environs. Wet removal of this aerosol will be low over the continent and the northerly parts of the Indian Ocean. Dry deposition on the ocean surface may be the only removal process for the regions devoid of precipitation. In the low latitude stratocumulus region, substantial cloud "processing" of air may occur which can alter the chemical and radiative properties of aerosols [Lelieveld and Heintzenberg, 1992].

The above features are very crucial for testing the direct effect of the sulfate-radiative cooling hypothesis [Charlson et al., 1992] which strongly relies on the following two assumptions:

(1) a fraction of the SO2 emitted is converted to new particles by gas-to-particle conversion; and

(2) sulfur, whether gaseous or particulate, is transported to regions far away from the source.

With respect to the first assumption we note that not all SO2 will be available to form new particles, as an appreciable fraction may be incorporated on soil dust, sea salt and smoke particles. Deposition on land and ocean surfaces may be another significant loss mechanism. If an appreciable fraction of the SO2 emission were to be cycled through these alternate pathways, the sulfate-cooling effect would be reduced. The second assumption is critically important, given the fact that lifetime of the sulfate particles ranges from a few days to a few weeks. Our current knowledge of this large scale spreading effect is largely based on chemistry-transport models [Langner and Rodhe, 1990].

Finally, an important dynamical feature of this region provides a critical test of the general circulation model (GCM) treatment of sulfate effects shown in Figure 6. The low level, northeasterly flow is confined largely to the region below the monsoonal low level inversion, at altitudes of about 1.5 to 2 km. Above this level the flow over most of India shifts from northeasterlies to westerlies. The pollutants that penetrate the inversion may be transported eastwards towards China, instead of to the Indian Ocean. Thus the large scale oceanic effect of sulfates emitted by India and the Far East may depend critically on the treatment of boundary layer processes in models.

The favorable surface flow will enable us to examine not only the transport of sulfur and sulfate aerosols into the remote Indian Ocean, but also quantify the rate of formation of new sulfate particles.

Figure 6. INDOEX: Sulfate visible optical depth calculations for a) January, and b) Marhce in Indian Ocean region. The calculations are based on the sulfate transport study of Pham et al., 1995.

Seasonal Change in the Direct Effect of Aerosols

Figure 6 illustrates a model simulation of sulfate aerosol visible optical depth. These optical depth calculations are based on the sulfate transport study of Pham et al., (1995), and are generally consistent with Advanced Very High Resolution Radiometer (AVHRR) retrieved aerosol optical depths. In January little sulfate is transported from the sub-continent. By April, however, there is significant transport of sulfate aerosol over the Indian Ocean. Thus the January to April time period offers a chance to observe the transition from a relatively clean environment to a fairly polluted one. Note that this transition could be observed either by performing measurements at a fixed island point in the Indian Ocean and monitoring through January to April, or by performing flights from the Indian sub-continent region to the Indian Ocean.

The direct effect of the increase in sulfates is to increase the scattering optical depth of the atmosphere (in visible wavelengths), which in turn leads to an enhancement in planetary albedo. Such estimates assume pure sulfate particles which do not absorb in visible wave-lengths, i.e., the single scattering albedo is unity. Contamination by black carbon (e.g., from biomass burning) and dust (from the Saudi Arabian and Rajasthan deserts) can alter this picture significantly, leading to absorbing aerosols. Only circumstantial evidence exists for the global extent of this direct effect. For example, long term (decadal scale) negative trends in solar radiation have been observed in urban regions, but not in remote or oceanic regions [IPCC, 1990].

Since the Arabian Sea is largely devoid of clouds during this period, measurements over this region provide an excellent opportunity to assess the degree to which the sulfate and other aerosol particles reduce the solar radiation reaching the sea surface. In addition, the study of Han et al. (1994) indicates a significant gradient in effective cloud drop size in this region, which is taken up later.

Source of Deep Convective/Cirrus Cloud Systems

The Indian Ocean contains a very large body of warm water, second in area only to the western Pacific warm pool. Sea surface temperatures (SSTs) averaged over the equatorial Indian Ocean (5šN to 5šS) are in the range of 301 to 303 K (Figure 7a). Since the deepest convection is triggered generally over such warm waters, this region along with the western Pacific plays an important role in the diabatic forcing of the tropical circulation (Figure 7b); region of high cloudiness enclosed by dark blue shading). The cross equatorial low level flow (Figure 5) into this regime of deep convection is associated with the ITCZ. The deep convection gives rise to the highly reflective thick cirrus-anvil cloudiness (Figure 7c) and contributes significantly to the vertical exchange of aerosols and trace gases between the surface and the upper troposphere including possibly the lower stratosphere.

Figure 7. a) Averaged sea surface temperature (SST), b) longwave, and c) shortwave cloud forcing for the Indian Ocean region.

The deep convective/cirrus cloud systems found in these regions reflect a significant amount of solar radiation back to space. For example, the monthly mean short-wave cloud forcing over the ITCZ can reach about -60 to -100 Wm-2 (Figure 7c). It is of fundamental importance to estimate the excess cloud absorption in these cloud systems. Also, in a matter of hours these deep convective systems can transport pollutants and aerosols (dust from Saudi Arabian and Rajasthan

deserts, black carbon from biomass burning) to the upper troposphere, where they can influence the cloud radiative properties in ways we can only speculate.

Cloud absorption and reflection depend on microphysical properties, and because these properties are known only from crude estimates from satellite, it is of great importance to have a knowledge of the aerosol and ice particle properties in this region.

Juxtaposition of Polluted and "Pristine" Air

The region south of the ITCZ (say south of 10šS) is essentially uninhabited, and could contain extremely low levels of short-lived pollutants such as reactive nitrogen compounds, sulfate, and ammonia. Very few measurements have been reported, but they support the assumption of pristine conditions [Moody et al., 1991; Clarke et al., 1989; Fishman et al., 1990]. Cloud condensation nuclei (CCN) south of the ITCZ circulation should arise only from natural sulfur sources, especially dimethylsulfide (DMS), and should be relatively sparse. The region north of the ITCZ (say north of the equator) is the source of air pollution arising from industry, home cooking and heating, transportation, agricultural waste burning, shifting cultivation, and emissions from animal waste.

Because of this juxtaposition of polluted and aerosol laden air to the north of the ITCZ with pristine air to the south, INDOEX offers an excellent opportunity to observe the role of ITCZ in the interhemispheric transport of trace gases and pollutants and to study how anthropogenic and natural aerosols alter cloud-radiative forcing. Of particular importance is the indirect radiative cooling effect of sulfates (which is pursued next) and the direct greenhouse effect of ozone which is described last in this chapter.

Ideal Region for Testing the Indirect Effect of Aerosols

The indirect effect arises through the ability of aerosols to act as CCN. Assuming unchanged liquid water content in clouds, the enhanced availability of CCN can increase the droplet number, their total area and thereby their reflectivity. This indirect effect can be as large as the clear-sky effect (Figure 1) [also see Jones et al., 1994; and Boucher and Lohmann, 1995]. In addition, by increasing CCN smaller cloud droplets are formed, which may reduce the efficiency of precipitation formation through a decrease in the droplet coalescence efficiency. The result is increased cloud thickness and lifetime [Albrecht, 1989] which in turn can alter cloud reflectivity [Pincus and Baker, 1994]. However, the uncertainties in the estimated effects are large [Han et al., 1994 and Boucher, 1995].

About 5š north and south of the ITCZ, marine stratocumulus clouds abound, instead of the cumulonimbus clouds in the ITCZ (Figure 5). In these shallow clouds, one might see clearly the effect of aerosols on clouds, i.e., sulfur and aerosol rich continental air to the north versus more pristine air in the region 5š south of the ITCZ.

In the upper troposphere, ice production will be modified by changes in aerosol composition and number density. This will be manifested through changes in ice particle concentration and size. This will in turn influence the radiative properties.

Convection and Ozone Formation

A preliminary trajectory study based on European Centre for Medium Range Weather Forecast (ECMWF) analyzed meteorological fields (described in Chapter 4) suggests that during the winter monsoon, pollution from the Indian subcontinent can be transported into the ITCZ in about a week. Such long range transport of air masses from the subtropics into the equatorial convection zone has also been suggested by CEPEX data [Kley et al., 1995]. The chemical lifetime of ozone over the tropical oceans is about one week, so that a considerable fraction of ozone formed from anthropogenic sources may reach the ITCZ. From here it can be transported into the cold upper-troposphere. This mechanism seems to enhance ozone's radiative forcing of climate. Furthermore, southward transport of relatively shortlived anthropogenic gases (such as nitrogen oxides, and non-methane hydrocarbons) may significantly affect oxidation processes over the tropical Indian Ocean. Since the tropical troposphere plays a major role in global hydroxyl (OH) formation, this may have consequences for the oxidation capacity of the atmosphere.