.gif)
The EXPeriment
for Regional Sources
and Sinks of Oxidants (EXPRESSO)
Contributed by:
Robert Delmas, Laboratoire d'Aerologie, Tolouse, France and Alex Guenther, NCAR-ACD, Boulder, USA
EXPRESSO was an international and multidisciplinary program, conducted by European, American, and African scientists, to quantify and better understand the processes controlling surface fluxes of photochemical precursors along a tropical forest to savanna gradient in central Africa, and to evaluate and improve understanding of the production and loss of tropospheric oxidants within this region.
This experiment associated ground-based, aircraft, and remote sensing measurements, allowing investigation of surface fluxes of energy and trace gases from both savanna and forest ecosystems on one hand, and photochemical and dynamical processes in the lower atmosphere on the other hand
The EXPRESSO domain extended from the savannas of the Central African Republic (CAR) in the north (8° N) to the tropical forests of the Republic of Congo (2° N) in the south. A research aircraft (the French Fokker 27-ARAT), instrumented for dynamics and chemistry measurements flew 11 missions out of Bangui, CAR, that covered a gradient from savanna to tropical forest. A 60-m walkup tower was installed at a nearly undisturbed tropical forest site. The tower provided access to the 45-m high canopy and the surface layer just above it. A detailed database of the extent of biomass burning in Central Africa during the EXPRESSO experiment was compiled using NOAA-AVHRR data (3 acquisitions per day). In addition, a detailed vegetation map was built for the EXPRESSO region with an evaluation of net primary production per type of ecosystem based on long term records of NDVI data. Finally a landscape VOC emission potential database was developed based on NOAA-AVHRR data, ground surveys of vegetation type, and a vegetation emission rate database
Figure 1. 3-D
modeling of CO distribution in Central Africa, in the boundary
layer (700 m altitude), on November 25, 1996, during the EXPRESSO
campaign. The biomass burning source is estimated from (i) vegetation
map and biomass density and (ii) from net primary production
estimates retrieved from NDVI data and modeling (Cautenet et
al., 1998).
|
Figure 1. 3-D modeling of CO distribution in Central Africa, in the boundary layer (700 m altitude), on November 25, 1996, during the EXPRESSO campaign. The biomass burning source is estimated from (i) vegetation map and biomass density and (ii) from net primary production estimates retrieved from NDVI data and modeling (Cautenet et al., 1998). |
Outline of the main results
The importance of the tropics on the chemical composition of the global atmosphere has been recognized for more than 15 years. The EXPRESSO field campaigns provide an original data set allowing to describe and better understand the atmospheric chemistry in the northern African tropics. It complements the experiments conducted since the middle of the 1980's such as GTE/ABLE 2A (1985) and 2B (1987) in the Amazon, DECAFE (1998, 1991) and STARE/SAFARI (1992) in Africa. The results gained from EXPRESSO allow us not only to document and parameterize biogenic and biomass burning sources, but also to study exchange processes of energy and trace compounds in the region of the ITCZ. The main results can be summarized as follows :
- The energy balance of savanna and forest ecosystems is primarily
determined by evaporation (latent heat flux) with Bowen ratios
of the order of 0.2 and 0.45 over forest and savanna respectively.
The kinetic energy budget in the boundary layer is dominated
by thermal (versus dynamic) production. This can be due to very
weak horizontal winds in the vicinity of the intertropical converge
zone (ITCZ), in spite of the strong roughness of both ecosystems.
The location of the experimental domain, at the interface between
savanna and forest ecosystems and in the middle of the ITCZ,
results in a great complexity of dynamical fields.Vertical exchanges
between the boundary layer and the free troposphere (monsoon
and Harmattan flows) result from a diversity of processes occurring
at various scales. They include (i) small scale processes such
as entrainment, penetrative dry convection, waves and wave breaking,
boundary layer clouds, and shear wind turbulence, (ii) mesocale
processes such as organized cloud coverage, local divergence-convergence
fields, penetrative cloudy convection and, (iii) synoptic scale
processes linked to subsidence and ascendance.
- A high density of fires was observed in savanna areas and,
as a consequence, a high level of pollution was observed at regional
scale. For instance regional concentration fields of CO and NOy
in the ranges 250'400 ppb and 4'10 ppb, respectively, and high
concentrations of benzene (0.5 ppb) and black carbon (10 µg
m-3) illustrate the overwhelming influence of biomass
burning on the chemical composition of the lower troposphere.
Biomass burning occurs in savanna areas where the average number
of fires detected every day in the "EXPRESSO window"
is about 5000 with corresponding burned areas of 20 000 km2/day.
However aircraft data taken over the forest and ground based
measurements taken at the forest site showed that biomass burning
pollution widely contaminates the forest atmosphere. Biomass
burning emissions were determined from (i) direct estimates of
burned areas and above ground biomass density derived from remote
sensing, and (ii) burning efficiency and emission factors. All
these parameters present rather large uncertainties. To better
constrain this dominant source, the redistribution of carbon
monoxide, considered as an inert tracer of biomass burning, has
been simulated using a 3-D non-hydrostatic mesoscale model coupled
with an inert tracer emission module. The objective was to compare
the CO distribution in the lower troposphere produced by the
model with vertical profiles of CO concentration obtained from
aircraft measurements, in order to assess the sensitivity to
the magnitude and location of sources in redistribution of an
inert tracer like CO and to constrain flux estimates. Estimated
fluxes have to be divided by a factor of 2 to retrieve CO concentration
fields consistent with aircraft observations ; estimates
of burned biomass in open fires is actually a difficult exercise.
- Biogenic hydrocarbon emissions were investigated on a variety
of scales along an ecological transect extending from the dry
woodland savanna of northern CAR to moist tropical forest in
the Congo. Above-canopy flux measurements at the forest site
characterized diurnal and seasonal flux variations that were
complemented by a regional spatial integration over the forest
and the savanna provided by flux measurements using an airborne
REA system. Mean isoprene fluxes observed with the aircraft system
were approximately 1200 µg(C) m-2 h-1
over the forest and 500 µg(C) m-2 h-1
over the savanna. In comparison, isoprene fluxes measured with
the tower based REA system over the forest ranged from nearly
zero at night to over 1000 µg(C) m-2 h-1
in the middle of the day. Seasonal variations in isoprene fluxes
observed at the tower site were greater than expected. A qualitative
emission characterization was obtained for over 200 dominant
vegetation species along the savanna to forest gradient and quantitative
rates of the major biogenic VOC were estimated for about 60 plant
species. Enclosure methods were also used to determine the relationships
between emissions and environmental conditions. These data were
used to develop a regional emission model which is in better
agreement with the aircraft measurements than are previous estimates
which were based on fluxes reported for a tropical forest in
the Amazon basin.
- Ozone and nitrogen oxide data allow study of photochemical characteristics of air masses in relation to ozone production efficiency (OPE). In the planetary boundary layer (PBL), high values of OPE are observed. concentrations increase with photochemical age of the air mass so that OPE is lower in the Harmattan layer. To complement the ozone budget of the PBL, ozone fluxes were measured from the aircraft flying at low altitude over the savanna and the forest. Average deposition velocities are equal to 0.7 and 1.5 cm s-1 respectively on both ecosystems.
Conclusion
The results gained from EXPRESSO allow us not only to document and parameterize biogenic and biomass burning sources, but also to study exchange processes of energy and trace compounds in the region of the ITCZ. Because of experimental constraints linked to the capabilities of the aircraft used, investigations were limited to 4 km altitude ; however model components can be evaluated using the experimental results. Based on the understanding gained from this work, improved sub-models will be developed and implemented in regional and global scale chemistry and transport models. This modeling effort will help us better understand the chemistry of the upper tropical troposphere which is essential to assess the oxidation capacity of the global atmosphere. The tropical troposphere is strongly affected by cloud convection which favors fast vertical transfers of ozone precursors (CO, NMHCs, and NOx, produced by the continental biosphere) within the rising branches of the Walker cells. These compounds drive O3production in the middle and upper troposphere. A significant impact on the stratosphere is also expected. Finally, in these regions, heterogeneous phase chemistry within convective clouds, assumed to be important, is almost unknown. Following EXPRESSO, future experimental projects in the tropics will focus on the impact of deep convection on the chemistry of the upper troposphere and lower stratosphere.