Figure 1. The climatological distribution of integrated tropospheric ozone determined from satellite data for the month of October. The stippled, hatched, and crosshatched areas indicate regions where more than 35 Dobson Units of ozone are present in the troposphere (from Fishman, 1994).

The Southern Tropical Atlantic Region Experiment (STARE) evolved from the designs for a collaborative project between Brazil and NASA, which had been proposed under the title TRansport and Atmospheric Chemistry near the Equator-Atlantic (TRACE-A) as part of NASA's Global Tropospheric Experiment (GTE). TRACE-A was designed to investigate by in situ measurements the extent, characteristics, and origin of the South Atlantic O₃ enhancement, and to assess the role of emissions from Africa and Brazil. Its primary goal was to determine the impact of fire emissions on scales that ranged from regional (~100 km) to quasi-global (>1000 km).

At the same time as TRACE-A was being developed, a team of African, European and American scientists came together in order to investigate the emissions from fires and soils in southern Africa, the meteorology over the subcontinent and the ecological impact of fires in the African savannas. This project, named Southern African Fire/Atmosphere Research Initiative (SAFARI), became, together with TRACE-A, the Southern Tropical Atlantic Region Experiment (STARE). SAFARI and TRACE-A complemented each other both in terms of scientific disciplines and of scales of interest. Thus, the combined missions provided a comprehensive multi-disciplinary experiment that spanned spatial scales from less than 1 km to more than 1000 km to provide insight into the potential impact of biomass burning on the atmospheric and terrestrial environment.

In spite of many political, administrative, logistical, technical, and meteorological obstacles, both teams, TRACE-A and SAFARI, managed to get into the field in the second half of 1992. TRACE-A involved some 200 scientists from the United States, Brazil, Congo, South Africa, and Great Britain. Its research was centered around the NASA DC-8 instrumented research aircraft. More than 150 scientists from 14 nations (Germany, South Africa, Zimbabwe, Zambia, Namibia, Swaziland, Congo, Brazil, Belgium, France, Great Britain, Canada, and the USA) participated in SAFARI, which involved several aircraft, ground-based operations, and remote sensing. STARE's interdisciplinary character was reflected by the involvement of research groups from atmospheric chemistry, meteorology, climatology, biogeochemistry, and fire ecology, as well as forestry, soil science, pasture science, and microbiology. Some key results from STARE will be presented in the following paragraphs. More information can be found in a special issue of the Journal of Geophysical Research (Volume 101, No. D19, 1996).

Fire behavior, fire ecology and ground-based emission measurements, including fire and soil emissions

A central objective in SAFARI was to conduct an integrated study of fire and its effects, including fire behavior and ecological responses, as well as the emissions from the fires and from soils impacted by burning. Energy release measurements and convection column monitoring showed that savanna fires, unlike boreal forest fires with their greater fuel loads, are unable to generate sufficiently high, sustained energy releases to produce convection columns above 3-4 km. Therefore, their emissions are confined to the lower troposphere, unless they become entrained by independently generated large-scale convective activity, e.g., after transport into the ITCZ region. Extensive measurements of fuel biomass, combustion factors, fire behavior, and emission factors were made in savannas, grasslands, and woodlands in South Africa and Zambia. As a result of this work, combined with previous studies in other savanna regions, we now have a very accurate and comprehensive inventory of emission factors from savanna fires, probably the best for any ecosystem (Cachier et al, 1996; Lacaux et al., 1996; Andreae, 1997; Koppmann et al., 1997). Scaling techniques developed under SAFARI made it possible to derive regional emission estimates bases on these emission factors and on fire detection by remote sensing. The results showed large differences to previous assessments, mostly due to much lower estimated amounts of biomass burned.

Studies of soil emissions of trace gases during SAFARI-92 concentrated on, NO, NO_x, and N₂O, and on trace carbon gases. Highest emission rates in the absence of burning were related to high soil total N content and N nitrification rate. Seasonal variations in NO emissions were ascribed to the effect of seasonally-varying rainfall on soil water contents (Levine et al., 1996). It is suggested that African savanna soils can provide a significant source of NO either with or without burning, depending on their moisture status and the seasonal timing of rainfall events. Soil fluxes of CO₂measured during SAFARI-92 in the semi-arid savanna of the Kruger National Park showed little influence of burning, but were also strongly affected by the availability of moisture (Zepp et al., 1996), increasing by an order of magnitude with heavy rain and maintaining elevated values for at least a week. In contrast, CO fluxes (which were considerably higher than those previously measured in southern African savannas) were insensitive to moisture changes, but responded positively to burning.

Airborne sampling of trace gases and aerosols

In addition to a focused study during September-October 1992, atmospheric measurements during STARE included enhanced and coordinated ozonesonde and rawinsonde launches from sites spanning the tropical Atlantic region. During this period, measurements were made using the NASA DC-8 aircraft, two Brazilian and five South African aircraft (with international science crews), as well as a series of ground-based stations on both continents.

TRACE-A's primary platform was the NASA DC-8 "Flying Laboratory", with a suite of in situ instruments that measured NO, NO_x, nonmethane hydrocarbons, H₂O₂, CH₃O₂H, CO, CH₄, N₂O, PAN, O₃, C₂Cl₄, PPN, HNO₃, some organic acids, and remote sensing capabilities for the measurement of ozone and aerosols above and below the airplane. As a reference point against which to study the impact of fire emissions, TRACE-A also characterized pristine air by flying as far south as possible off the east coast of South America to find an airmass that had not recently been influenced by continental emissions. During the southernmost portions of this flight, NO_x < 18 ppt, CO < 65 ppb, CH₃CL < 600 ppt, and C₂H₂ < 45 ppt were measured (Talbot et al., 1996) indicating that the air had likely not been influenced by continental emissions for more than 10 days. A synthesis of the measurements from these instruments was then used to characterize the emissions from the continents, to understand the photochemical processes taking place in the atmosphere downwind from the emissions, and eventually to glean insight into how these emissions and subsequent photochemistry impacted the global scale aspect of the composition of the troposphere.

A significant component of TRACE-A included a suite of investigations fielded by the Brazilian Space Agency, INPE, in collaboration with other Brazilian agencies and universities. The Brazilian investigations consisted of two aircraft instrumented for measurements of trace gases (CO, O₃, N₂O, CH₄, and CO₂), radon, and aerosols (black carbon), respectively. These aircraft obtained vertical profiles of these species in addition to defining spatial gradients within Brazil during survey and transit flights (Kirchoff and Alvalá, 1996). Surface measurements of these species were made within the region generally affected by biomass burning at three ground sites.

The DC-8 flights in Brazil were designed to accomplish the primary objectives of characterizing the composition of air exiting the continent that had been influenced by local sources, characterizing the composition of air that had been recently impacted by continental emissions, and also obtain some measurements of air very near the source regions (i.e., fires). An additional "objective of opportunity" was to make a measurement of air that had been recently transported from the boundary layer to the upper troposphere by convection. One of the flights successfully captured this opportunity and found CO mixing ratios enhanced by more than a factor of three (200-300 ppb vs. 90 ppb) at altitudes of 9-12 km and clear signals of lightning-generated NO_x. It has been shown that the combination of convection and subsequent transport at high altitudes is important for contributing to the high ozone concentrations found over the southern tropical Atlantic (Pickering et al., 1996).

The prolonged 1991/92 drought in southern Africa, called by many the worst drought of the century, resulted in less biomass loading throughout the region and, therefore, probably reduced the amount of burning relative to what would have taken place during a more representative year. In Kruger Park, South Africa, the focal point of the SAFARI campaign, vegetation was so sparse in most places that it was questionable for some time if the planned experimental burns even could take place. But eventually a region in the southern part of the park was found that was suitable for burning, and a series of experimental fires was conducted. These fires were investigated by three of the SAFARI aircraft: a DC-3, a Cessna 182, and a Bell helicopter. All had been equipped with instruments for trace gas and aerosol measurement and sampling. The DC-3 payload included instruments for the determination of O₃, NO, NO_x, NO_y, CO₂, CH₄, NMHC, and aerosols. From the flights into the plumes of the experimental fires as well as numerous fires of opportunity came a comprehensive characterization of the emission characteristics of savanna fires.

Following the plume flights, the DC-3 undertook two "grand tours" of southern Africa, to map the impact of biomass smoke over the subcontinent, spanning the countries of South Africa, Zimbabwe, Zambia, Namibia, Swaziland, Angola, and Botswana. This work was complemented by detailed O₃ studies over Namibia by a Learjet aircraft. The NASA DC-8 traveled to northern Zambia (~1300 km) to find a region of widespread burning. Charred expanses of burnt area were observed and were indicative of the widespread nature of the recently burned areas throughout the study area. In these polluted regions, high concentrations of many trace gases were encountered up to altitudes of ~3000 m. O₃ generally ranged from 60-100 ppb; NO, 0.15—0.20 ppb; NO₂, 1.1—1.4 ppb; PAN, ~4 ppb; H₂O₂, ~5 ppb; HNO₃, ~1 ppb; HCOOH, 7—8 ppb; and CH₃OOH, 0.7-0.8 ppb (for further details, see Gregory et al., 1996; Harris et al., 1996; Heikes ept., 1996, Jury et al., 1996; Singh et al., 1996; Smyth et al., 1996; Talbot et al., 1996; Thompson et al., 1996; Zenker et al., 1996). Aerosols, hydrocarbons, CO, CH₄, N₂O, and CO₂, were also elevated well above their background concentrations (Anderson et al., 1996; Blake et al., 1996; Le Canut et al., 1996; Maenhaut et al., 1996). Although O₃ concentrations were high in the same region as the most intense smoke, higher mixing ratios were observed above the boundary layer (Figure 2.2). In a layer generally situated between 4-7 km, the UV-DIAL measured mixing ratios were often in the 80-120 ppb range (Browell et al., 1996). In addition to biomass burning, large NO emissions from savanna soils following rain showers were found by the DC-3 during a flight over Namibia (Harris et al., 1996).

Figure 2. Vertical distribution of ozone in the tropics. There is no biomass burning influence evident over the equatorial Pacific, some influence of transport off the west coast of South America, and strong O₃ enrichment by photochemical production in biomas burning plumes over Brazil and the Congo. The STARE data (shaded) are the highest regional ozone levels recorded anywhere in the tropics and subtropics to date.

Pg. 2 >>