Abstract: Soil heat flux was calculated by using methods such as heat flux plate measurement and calorimetry (PlateCal), thermal diffusion equation and correction (TDEC), harmonics, conduction-convection, amplitude, and phase to examine the observations of "Microclimate and Evaporation Experiment in Inner Mongolia". The authors compared these six soil heat flux calculation approaches and researched their applicability to various surface conditions. Moreover, the authors analyzed their impacts on surface energy balance closure. In general, the PlateCal method was shown to accurately calculate soil heat fluxes at depth of 2 cm with a mean square error of 6.9 W/m². In different surface conditions, this method is the best approach for calculating soil heat fluxes in dry and rainy conditions, with mean square errors of 14.0 W/m² and 21.4 W/m², respectively. The harmonic method is the best approach for calculating soil heat fluxes in wet conditions, with a mean square error of 30.1 W/m². The largest difference in surface soil heat fluxes calculated by the six methods was 178.6 W/m². Large differences in surface soil heat flux exceeding 25 W/m² accounted for 96.3% of the sample. The six soil heat flux calculation approaches can obviously affect the energy closure rate; the energy closure rate increases with an increase in turbulence intensity.