PhD Thesis
Turbulent exchange of ozone and nitrogen oxides between an Amazonian rain forest and the atmosphere
Udo Rummel (11/2005-11/2005)
Support: Thomas Foken
Betreuer im Max-Planck-Institut für Chemie: Prof. Dr. F. X. Meixner
Amazonia, the world’s largest tropical rain forest area is facing rapid development, mainly caused by slash and burn activities. The change in land use, primarily to agricultural and pasture areas, has sustainable influence on the atmospheric input and the deposition of constituents like ozone (O3) and nitrogen oxides (NOx), which are of high relevance for tropospheric chemistry. To asses the effect of land use change on tropospheric chemistry, a good knowledge of the exchange of those trace gases between the primary rain forest ecosystem and the atmosphere is necessary. So far, experimental information from tower based canopy-scale and leaf-scale measurements focusing on exchange processes is very limited. Within the framework of the LBA-EUSTACH project 1999 two experiments were carried out in Rondônia, southwest Amazonia, to estimate the exchange of ozone and nitrogen oxides between a tropical rain forest ecosystem and the atmosphere during the wet and the dry season, respectively.
Ozone deposition was determined by eddy covariance measurements above the canopy. The data obtained during the end of the regional wet season confirm the results of the only previous study reporting on directly measured O3 fluxes above the Amazonian rain forest. Mean daytime maxima of -11.0 nmol m-2 s-1 and 2.3 cm s-1 for ozone flux and deposition velocity, respectively, show the rain forest to be an effective sink for ozone during the wet season. At the end of the dry season, under conditions of high atmospheric humidity deficit, the ozone uptake by the forest canopy was significantly reduced. A consequence of this strongly reduced uptake was a substantial in-canopy O3 storage during day, which was removed in the first half of the night, by considerable non-stomatal deposition and chemical destruction.
Ozone deposition was simultaneously determined at an old pasture site which was deforested 22 years before the LBA-EUSTACH experiment. The measurements at this site showed an ozone deposition velocity ~35% and ~25% lower than the rain forest values for the end of the wet and dry season, respectively. Since cattle pastures represent the largest part of converted forest land in Rondônia, this difference may represent the effect of deforestation on the regional O3 surface sink. Based on land cover information provided by LANDSAT images, the current regional O3 deposition average for central Rondônia was estimated to be ~85% of the original sink provided by the native rain forest cover.
NO soil emissions were determined by an eddy covariance system which was positioned within the trunk space. Nighttime measurements resulted in mean values from 3.5 ng N m-2 s-1 to 4.8 ng N m-2 s-1, in good agreement with emission fluxes obtained by concomitant dynamic soil chamber measurements.
A further aim of the experiment was to characterize the turbulence structure throughout the canopy during two intensive measuring periods. Detailed analysis of high frequency time series of several scalar quantities above and within the canopy revealed, during daytime, the frequent appearance of ramp pattern, the "finger print" of coherent turbulent structures. This enabled (i) to estimate a mean residence time of air within the part of the canopy which is directly coupled to the atmosphere above by these short, extreme, exchange events, and (ii) to determine ozone fluxes by a surface renewal model based on coherent air motion.
To assess the relevance of in-canopy processes as turbulent transport, uptake by vegetation, soil deposition, and chemical reactions to the ecosystem exchange of ozone and nitrogen oxides, their characteristic time scales where analyzed.
For the first time NO2 profiles were measured within a tropical rain forest. By combining these results with all available wet season data on leaf level exchange in a stationary budget approach, a reduction of soil-emitted NOx by vegetation up to 25% was obtained. This value is considerably smaller than that obtained by previous model studies. Direct comparison to the NOx budget of the old cattle pasture indicated the primary rain forest ecosystem to be a higher NOx source, and suggests therefore, that deforestation is reducing the biogenic NOx emission in southwest Amazonia on a long term basis, if no fertilizer is used.