Amazon region is known worldwide for the availability of water and for the number of
ecosystems such as dense forests land, flooded forests, flood plains, igapós, open and closed
fields. The important role that the Amazon rainforest plays in the energy and mass exchange
with the atmosphere and the implication of these changes in the climate at a local, regional
and global scale is a fact difficult to be answered. Since such changes are controlled by
atmospheric turbulence, the understanding of turbulent flow in the different atmosphere
layers, in and above the forest canopy and above water surfaces in the Amazon, becomes
necessary. In this work, the vertical variability of the wind velocity, the turbulence statistical
moments, the sensible heat flux (H) and the turbulent kinetic energy dissipation rate (ε) in
different experimental sites in and above Amazon forest were studied. In addition, a
comparison was made between turbulent flows above one of these sites (a rough surface) and
above a lake, which is also located in the Amazon. Regarding the vertical wind profile, data
collected from six towers of different experimental sites were analyzed, aiming to observe the
general characteristics of the behavior of said profile, as well as to test the ability of simplified
models to reproduce such characteristics. In general, the profile below the canopy is strongly
affected by the forest structure. From the soil up to 0.65h (where h = 35 m is the average
height of the forest canopy), the vertical wind profile is approximately constant with height
and has very low values, less than 1,0 ms -1. Above 0.65h up to 2.25h the wind speed
increases with height. As for the models used, both the Yi and Souza models were able to
reproduce satisfactorily the wind profile for the different experimental sites. In relation to this
last model, it was still possible to reduce the amount of input variables required to simulate
the vertical wind speed profile without prejudice to the aforementioned ability. Differently
from the wind profile, the other analyzes performed in this work were based on data from
only two experimental sites. These data were collected by bi and three - dimensional sonic
anemometers arranged from near the forest floor to about 80 m high. Comparing the results of
the two sites studied in this study with other two (investigated by other authors in published
works) also located in Amazonia, these sites showed significant differences in the efficiency
of absorbing momentum of the atmosphere, probably due to small differences in the forest
structure of each site. The behavior of the turbulence statistical moments showed that eddies
generated above the forest canopy hardly penetrate the region below 0.5h, , and that depth can be more easily reached during strong wind conditions. Likewise, H values were higher during
this condition in both during the daytime and nighttime. Another important observation
disappears from the H profile is that is not constant with the height, compromising the
validity of the Monin-Obukhov’s theory similarity. In addition, it has been noted that the
behavior of the H diurnal cycle is quite complex at certain times, changing from positive to
negative within the daytime period. The ε diurnal cycle shows the same behavior at all
heights, influenced by the solar radiation diurnal cycle. The highest values of ε were also
found during the strong winds performance, with a maximum close to the forest canopy top.
Comparing the turbulence characteristics above the forest with those observed above the lake,
it was verified that in general the air turbulence intensity above the forest was higher than
above the lake during the daytime, due to the high efficiency of the forest in absorbing
momentum of the turbulent flow. During the nighttime the situation was reversed, with
greater air turbulence intensity above the lake, except in some periods in which intermittent
turbulence bursts occur above the forest. The horizontal (Λu) and vertical (Λw) eddies scales
calculated during the daytime period were higher above forest than above the lake, and the
vertical length scale (Lw) was also larger over the forest, but the horizontal length scale (Lu)
was higher above lake. The composites of the vertical velocity power spectra obtained for the
daytime and nighttime periods for each site showed canonical behavior, with a well-defined
inertial subdomain region, except in the nighttime period above the forest. Finally, shear
production was the dominant term of the turbulent kinetic energy (TKE) budget equation
during the daytime period at both sites. All terms calculated in this work at night showed
values close to zero over the lake, indicating that the terms that could not be calculated, such
as the TKE advection, may have contributed to maintenance of turbulence overnight at this
site.
|
