In the past decade, hurricane model forecast skill has improved as a result of better physical parameterizations and data assimilation schemes. Even with these improvements, the models still experience some lack of skill when forecasting specific aspects such as cyclogenesis, precipitation, surface winds, etc. Comparisons between model-predicted hurricane track and intensity, and official forecast have traditionally been used as the main verification of the performance of the models. This method does not provide an accurate means of assessing the model performance because it is evaluating the general performance of the model and not its dynamical structure. The objective of this study is to examine existing observational data sets, other than track and intensity, to evaluate the model's performance and find better-quality indicators of forecast skill. A comparison between observations from different platforms (e.g. aircraft, satellite, weather stations, sondes, ships) and output from the Geophysical Fluid Dynamics Laboratory (GFDL) model has been performed. The inner grid wind data at low levels are being brought to H*Wind, a global tropical cyclone observational analysis system developed at the Hurricane Research Division of NOAA's AOML, to make the comparisons using error covariance estimation. Data from Hurricane Ivan (2004) is used to provide an insight of the model efficiency for models currently in use.