Abstract: The deficiency of the eddy covariance method is an important cause of surface energy imbalance. This article conducts research on the large eddy average method and Durand method proposed in the literature to overcome the deficiency of eddy current correlation method and achieve surface energy balance. The similarities and differences, comprehensive effects, and improvement methods of the two methods were studied through five groups of test during the daytime. The results showed that: (1) The closure level of surface energy is related to the length of the sequence; (2) the original Energy Balance Ratio (EBR) was 0.80, and the Energy Residual (Res) was 64.9W m-2^{; (}3) After considering the large eddy averaging method, the EBR increased to 0.98 and the Res decreased to 14.5W m-2^{, reaching a} balance; (4) After considering the Durand method, the EBR increased to 0.95 and the Res decreased to 25.1W m^-2, which was less effective than the large eddy average method; (5) After considering both the large eddy averaging method and the Durand method, the EBR increased to 1.17 and the Res decreased to -35.1W m^-2, resulting in severe over closure. Analysis reveals that the cause of severe over closure is due to the Durand method"s excessive accumulation of gas expansion and work done; (6) When not considering the sensible heat additional term of the Durand method, only considering its latent heat additional term, and simultaneously considering the large eddy average method, the result can realize an ideal surface energy closure state, which EBR achieves to 1.0 and Res is reduced to 8.7W m-2^{, only account}ing for 3.2% of the available energy. Analysis shows that the large eddy averaging method is the absolute dominant factor in achieving surface energy closure.