Abstract: The error of sea surface wind derived from models is vital for hurricane wind hindcast. Currently, reconstructing sea surface wind field in the boundary layer through mathematic models is a fast and commonly used method. Taking “IKE”, the 8th Atlantic hurricane that invaded the Gulf of Mexico in 2008, as an example, this study compared the surface wind field of “IKE” derived from 8 kinds of widely used mathmatic models, and the results showed that different models performance diversely. The Jelesnianski-1 model can effectively represent the spatio-temporal characteristics of Typhoon Ike's wind field, in addition, models (e.g. Holland model and Miller model) with variable coefficients also showed high applicability, the results of which can be in line with the actual typhoon cases by adjusting the value of coefficients, Overall, sea surface wind of hurricane reconstructed by mathmatic models can basically reflect the characteristics of the wind field in the core area of the hurricane, but failed in giving accurate background wind field in the outer area; while the analysis or reanalysis wind field data can provide precise wind field in the outer area of hurricane. Therefore, integration of wind field data derived from mathmatic models and analysis or reanalysis wind field can significantly reduce the error of sea surface wind simulation of hurricane.