Development of advanced spectral wind wave models has been the subject of comprehensive researches which has led to reliable wave predictions for assessing the impact of waves on the natural environment, coastal protection, offshore structures and harbor over the past two decades. On the other hand, the Geographical Information Systems (GIS) are developed for working with geographical data which as standard tools are able to manage and represent such data. In the present work, linking the third-generation wind wave model of SWAN and GIS has been implemented to ease data management and representation of inputs and outputs of the model that are geo-reference data. Then the effect of computational time step has been investigated using available data for Urmia Lake situated in the Northwest of Iran. As SWAN has the ability to perform the options of 1st, 2nd and 3rd generation modeling, these options have been used to evaluate results obtained by applying these models. Results indicate that the 3rd generation wave model is more sensitive to higher computational time steps than other modes of wave generation models. However, the convergence behavior of the third-generation model is much faster than that of other models.