[A comparison of various correction methods for calculating carbon flux above tropical seasonal rainforest]

In this paper, various correction methods were compared, based on the observation data of the CO2 and water fluxes above the canopy of a tropical seasonal rainforest in Xishuangbanna during 1-9 March, 2003. The results showed that it was necessary to correct the observation data when calculating the carbon flux, though various correction methods contributed differently to the results. For daytime data, WPL made the most important contribution, while sonic temperature correction did the least. t test indicated that no significant difference was found between the carbon fluxes corrected by WPL and the standard values. Cluster analysis showed that triple-rotation correction was helpful to improve the precision of the data. For nighttime data, it was the triple-rotation correction that made the most important contribution, and sonic temperature correction also did the least. There existed significant differences between the carbon fluxes corrected by WPL, planar-fit correction and their combination and the standard values, while the differences between the carbon fluxes corrected by double-or triple-rotation and the standard values were not significant. Cluster analysis showed that the data corrected orderly by coordinate axes-rotated correction and WPL would be propitious to improve the precision. The results would provide a basis for calculating the fluxes within tropical rainforest in the future, and supply reference to the fluxes calculation for other areas.