Abstract: The lack of unanimous opinion in spectral correction of energy and carbon dioxide eddy covariance flux error caused the difficulties not only in intercomparison of results derived form the different experiments, but also in synthesis data for global models. Two kinds of spectral methods used to correct sensor error were introduced. The influence of two-dimension and triple–dimension air current on CO₂ flux was discussed. Meanwhile, Error encountered in processing data and problems of measurement CO₂ flux in nighttime were investigated.