Use of the inertial dissipation method for calculating hourly variability of the turbulent fluxes over Amazonian forest

The well-known Inertial Dissipation Method (IDM) to is used to calculate the turbulent fluxes of the heat and momentum above the Amazonian Forest. Assuming equilibrium between production and dissipation of turbulent energy and the validity of of similarity relationships of the Monin-Obukhov surface layer parameterization, the power spectra of one wind velocity component and temperature in the inertial subrange is sufficient to compute the fluxes. So, fast response data measured in short time intervals (a few minutes) is sufficient to get reliable estimates of the fluxes. This is in contrast to the other methods such as eddy-correlation and flux-variance, which require data for time intervals greater than 20 min to give reliable flux estimates. Even though, IDM cannot be used in certain conditions such as free convection, it permits a better characterization of the small scale temporal variability of the fluxes. At the same time, it is able to detect flux fluctuations associated with mesoscale, individual eddie motions and other phenomena that occur in the tropical boundary-layer.

In this work we examine how the temporal variability of the fluxes increases as the integration time decreases. We also investigate what kind of processes determine the flux variability during the 1 hour time scale, for a wide range of stability conditions. This study, was done using the turbulent data measured above the Amazon Forest canopy, on a micrometeorological tower built in the Rebio-Jaru Biological Reserve (10o 04' S; 61o 56' W), in northwestern state of Rondonia, Brazil, during LBA (Large Scale Biosphere Atmosphere Experiment in Amazonia) 1999 wet season campaign. A Campbell sonic anemometer and a fast response thermometer placed at 67m above the ground provide the turbulent data. The data was sampled at 16Hz. Preliminary results show that the IDM algorithm for unstable conditions presents surprisingly good convergence index in the time interval from midday to early night but not during the early morning. This could be due to the peculiar multiscale nature of vertical transport processes in the tropical boundary layer (1).

(1) Williams, A.G.; Kraus, H.; Hacker, J.M. Journal of the Atmospheric Sciences v.53, no 8, pp. 1187-1202, 15 april 1996.