Three surface layer schemes (MYNN, ETA, and MM5) in the Weather Research and Forecasting (WRF 4.1) model were selected to conduct simulation experiments for tropical cyclone (TC) Hato (2017). The influence of different surface layer schemes on TC intensity was compared. Although the planetary boundary layer and land surface schemes were the same, surface layer schemes minimally affected TC tracks but significantly affected TC intensity. The highest correlation coefficient and smallest bias and root mean square error between the simulated TC intensity in the ETA and observations. The TC intensity simulated using ETA was the strongest and closest to the observations. The surface latent heat fluxes simulated using ETA at the time of the TC landing were 1200–1500 W/m2, and is 800–1200 W/m2 by MM5 and MYNN, respectively. The surface sensible heat flux simulated using ETA was 400–800 W/m2 and that simulated using MM5 and MYNN was 200–400 W/m2. The equivalent potential temperature and radial wind simulated using ETA were significantly higher than those simulated using MM5 and MYNN. The larger the energy transmitted from the surface to the tropical cyclone, the stronger the TC. The dynamic structural characteristics were consistent with the tropical cyclone intensity. The ETA indicated the strongest unstable planetary boundary layer structure, stronger convergence at the lower layer, divergence at the upper layer, vertical transport, and absolute angular momentum, and induced the formation of the strongest TC. The comparison results revealed that the surface fluxes from different surface layer schemes can affect the dynamic and thermal structures of the planetary boundary layer owing to the different stability functions. The surface fluxes were also affected by the temperature and wind in the planetary boundary layer. This interaction between the surface layer and planetary boundary layer affects the eye structure and secondary circulation of the TC and TC intensity.