Final calculations

Turbulent fluxes are calculated from covariances of the despiked and detrended time series taking into account the delay time of signals determined during the calibration procedure outlined in section . Periods with incomplete sampling over the entire hourly period are not used for flux calculation. Raw covariances are converted into physical quantities using equations -.

To determine the quality of the measurement, and to classify the data Foken and Wichura (1996) propose three tests: the instationarity test, the correlation test and the integral turbulent characteristics test. In our case the instationarity test of the hourly flux values is applied to qualify the data calculated. The quality flags (e.g., the percentage of instationarity) can be used in the determination of the net carbon balance of the region in such a way that flux values beyond a specific threshold instationarity are not used, resulting in less random error in the measurement. However, it should be noted that the effect of random error in the determination of the net carbon budget of the region is not severe if the time series available is long enough (Moncrieff et al., 1996).

Since the fluctuations of HO and CO measured by the LI-COR 6262 are expressed in terms of mole fraction relative to dry air it is not necessary to perform the corrections for variations of the density of air (Webb et al., 1980; Grelle and Lindroth, 1996; Berger et al., 2001).

Additionally, Obukhov length is calculated from the turbulent characteristics and stored to provide stability information neccessary for the spectral corrections and for the flux source area analysis.