The Averaging Period for Eddy covariance method-the Chi-Lan Mountain Site

Autor: Ji-Yu Lin, 林繼于
Rok vydání: 2010
Druh dokumentu: 學位論文 ; thesis
Popis: 98
The eddy covariance method has been used to evaluate the fluxes of scalars and energy between the ecosystem and atmosphere. It bases on the Reynolds decomposition theory to separate the change of scalars and energy into the average and variance components. Through the time average procedure on the instantaneous measured raw data, fluxes carried by atmospheric turbulence can be estimated by the covariance of deviations of mean scalars and wind. However, if the chosen averaging period is too short, fluxes carried by low-frequency eddy might be lost resulting and underestimation of fluxes. On the other hand, if the chosen averaging period is too long, the estimatited fluxes might be influenced by non-stationary flow and violate the assumptions of Reynolds decomposition theory. In this study, the cospectrum and ogive function of scalars and energy fluxes were calculated and a suitable averaging period were examined for the Chi-Lan Mountain flux site. During the day, the strong convection resulted the lower-frequency eddies. The best averging period to estimted carbon dioxide, water vapor and sensible heat flux is about 60 minutes. However, the ogive function parameters showed that the ogive value of averaging period of 30 minutes suggested by most literatures was convergence. Comparing to the period of 30min, carbon dioxide, water vapor and sensible heat flux is more than 0.0%, 5.0% and 2.0% by using an averaging period of 60 min. And it’s helpful to improve the closure of energy budget. During the night time, the flow was mainly caused by shear stress and the resulting turbulence size was smaller. However, there existed mesoscale flow probably induced by the topography around the flux tower. The timescales for flux transport were variable and may be not less than the daytime period. Since the intensity of fluxes during the night time were low, the influence of flux estimations using different averaging period is very large. Therefore, choicing the averaging period will be very important in order to avoid additional uncertainty of the flux estimations during the night time. In addition, the cospectrum of sensible heat flux showed that the direction of the flux in different scales were different. The mechanism for this phenomeenon is not clear yet, further measurement and confirmation is needed to verify it.
Databáze: Networked Digital Library of Theses & Dissertations