4.1.1 Primary Rain Forest
In Rummel et al. [2005a], surface deposition, the main sink for tropospheric ozone was determined at different time and spatial scales. Tower based eddy covariance flux and mixing ratio profile measurements at the rain forest site, complemented by leaf level information form branch cuvette measurements revealed new information about seasonal and diel behavior of O3 deposition to that tropical ecosystem type.
The eddy covariance measurements at the end of the wet season (LBA-EUSTACH 1) showed the tropical rain forest in the south western part of the Amazon basin to be an effective O3 sink (Figure 6 (a)). Magnitude and diel variation of O3 flux and deposition velocity (mean daytime maxima:
-11.0 nmol m-2 s-1 and 2.3 cm s-1) are comparable to the wet season O3 deposition found for a tropical rain forest in central Amazonia by Fan et al. [1990].
0000 0400 0800 1200 1600 2000 2400
-20 -15 -10 -5 0 5 10
(a)
O3 flux (nmol m-2s-1)
time (LT)
0000 0400 0800 1200 1600 2000 2400
-20 -15 -10 -5 0 5 10
(b)
O3 flux (nmol m-2s-1)
time (LT)
Figure 6: Diel variation of mean (median) O3 budget components for EUSTACH 1 (a) and LBA-EUSTACH 2 (b): O3 flux measured at 53 m (black squares), O3 storage within the column 0-53 m (open triangles), net “chemical” flux (chemical reactions of the NO-NO2-O3 triad: black dotted line), O3 flux at the forest soil surface (gray circles), and O3 flux corrected for storage (black line; gray area indicates its inter quartile range). Figure taken from Rummel et al.
[2005a]
The results of LBA-EUSTACH 2 provide the very first information about the diel course of O3 deposition to a tropical rain forest under dry season conditions, partly influenced by substantial
biomass burning activities (Figure 6 (b)). In contrast to the regional scale information from former integral boundary layer budget estimates from central Amazonia, e.g. by Gregory et al. [1988], the tower based eddy covariance measurements enabled the investigation of mechanisms which regulate surface deposition of O3. The results indicate, that the daytime O3 deposition to the forest at the end of the local dry season was influenced by specific humidity deficit SHD and consequently controlled by stomatal aperture. Days characterized by moderate SHD values showed rather higher daytime O3 fluxes than those days with extreme specific humidity deficits, although systematically higher O3
mixing ratios occurred during the latter period. Therefore, the concurrence of high specific humidity deficits and O3 mixing ratios partly up to 80 ppb above the canopy might be also caused by the limited daytime O3 uptake of the forest (under extremely dry conditions), beside enhanced atmospheric O3 production as a consequence of biomass burning.
The consequence of the strongly reduced uptake during the dry periods of LBA-EUSTACH 2 was a substantial in-canopy O3 storage during day (Figure 6 (b)). The stored O3 was finally removed in the first half of the night by considerable non-stomatal deposition, i.e. to vegetation surfaces like leaf cuticle and bark.
The occurrence of cuticular O3 deposition was supported by branch cuvette measurements at a tropical deciduous tree species (Hymenaea courbaril L.) from the nearby IBAMA camp site. But the inferred mean cuticula resistance for O3 of ~ 4000 s m-1 to 5000 s m-1, derived through the application of a leaf resistance model according to Baldocchi et al. [1988], might be to high to explain the fast O3 depletion after sunset solely by dry deposition on plant surfaces. This makes additional chemical sinks likely.
The second remarkable result is the almost identical relative diurnal course of the mean O3 deposition to a branch of one tree species inferred from the cuvette measurements and the canopy scale O3 flux obtained by the eddy covariance measurements. This clearly shows, that during daytime the O3 deposition to the rain forest is dominated by the stomatal uptake of the crown species, which as a functional group are similar in their stomatal behavior [see Roberts et al., 1990; 1993; McWilliam et al., 1996].
4.1.2 The Impact of Deforestation
The O3 fluxes obtained at the primary rain forest site (RBJ) [Rummel et al., 2005a] were compared to O3 deposition measurements simultaneously performed at the 22 years old cattle pasture (FNS) [Kirkman et al., 2002]. The results showed a ~35% lower mean deposition velocity at FNS at the end of the wet season. At the end of the dry season the mean deposition velocity at the pasture site was
only ~25% lower compared to the forest. This smaller difference is mainly a consequence of enhanced canopy resistances at the forest as reaction on the high specific humidity deficits in the dry season.
Assuming both experiments to be representative for two equally long seasons (6 months each, which may be reasonable for Rondônia with respect to the monthly average SHD values [see Culf et al., 1996]) after deforestation the regional O3 surface sink would be about 70% of the original value for the extreme case of a total rain forest to pasture conversion. Since cattle pastures represent the largest part of converted forest land in Rondônia [Fearnside, 1980; Roberts et al., 2002], this scenario might not be unrealistic.
For an estimate of the current state and the recent history in central Rondônia a simple up scaling approach was made in Rummel et al. [2005a] based on the LBA-EUSTACH results and the land cover information provided by the analysis of Roberts et al. [2002] for the two LANDSAT scenes P231, R67 (Ji-Paraná) and P231, R68 (Luiza).
0 20 40 60 80 100
1984 1986 1988 1990 1992 1994 1996 1998 2000 0
20 40 60 80 100
(b)
land cover (%)
year
^ v (% of original O deposition)d3
0 20 40 60 80 100
1978 1986 1988 1990 1992 1994 1996 1998 2000 0
20 40 60 80 100
(a)
land cover (%)
year
^ v (% of original O deposition)d3
Figure 7: Temporal evolution of the areal percentage of the three major land-cover classes (primary rain forest: black column, pasture: white column, and secondary growth: gray column) of two LANDSAT scenes: (a) Ji-Paraná (~26500 km2) and (b) Luiza (~30000 km2) in central Rondônia [Roberts et al., 2002]. The black line indicates the corresponding change of the areal O3 deposition average in relation to the original forest cover. Figure taken from Rummel et al. [2005a]
Figure 7 displays the land cover history of the two scenes from 1978 and 1986 to 1999 for Ji-Paraná and Luiza, respectively [Roberts et al., 2002]. Both areas show a similarly reduction of the tropical rain forest cover from 82% and 79% in the year 1986 to 54% and 50% in 1999. The calculated reduction of the area averaged O3 deposition for the same time period is from 93% to 86% and from 92% to 84% of the original tropical rain forest value.