In order to obtain more accurate winter precipitation data, this paper aims at the correction and error calculation of the influence of near-surface horizontal wind during snowfall measurement using PARSIVEL2 (Particle Size and Velocity). The revised results show that under certain wind speeds, the influence of ignoring wind will cause significant underestimation of the diameter of large particles, while for particles in the same size, larger wind speed means the underestimation of particle diameter during calculation is more obvious. When the wind speed does not exceed 2m·s-1, the calculation error of the falling speed of the snowfall particles is about 3%, and the calculation error of the diameter is within 7%. In the analysis of the real snowflake spectrum obtained during a snowfall in Nanjing on January 4th, 2018, it can be seen that ignoring the influence of wind will lead to the shift of the peak of the snowflake spectrum and the narrowing of the spectrum, which will result in the overestimation of concentration of small particles and underestimation of the concentration of large particles, which in turn affect the calculation of microphysical quantities. Specifically, the radar reflectivity factor Z and the snowfall intensity I are underestimated, and the actual value of the Z-I relationship fitting coefficient a is greater than the calculated value, and b is smaller. However, when the wind speed is larger, the flow near the ground is more complicated, and the vertical turbulent motion cannot be ignored. This correction method is likely to be no longer applicable.It is recommended to add a windbreaker in future observations or to add corrections in subsequent data processing to eliminate the impact of wind on snowfall measurements.