Satellite observations provide valuable measurements for NWP models, particularly in the locations where conventional observations are sparse or inaccessible. One such measurement is the ocean surface wind velocity, available at wide swath and with high resolution and accuracy from NASA's QuikScat satellite. One of the important application of this observation can be its assimilation in numerical weather prediction models for improvement of the prediction of track and intensity of tropical cyclones. However, the fact that these data are available on just one level i.e. ocean surface level, poses a serious limitation as far as their assimilation in the models is concerned. Our study indicates that a direct assimilation of these winds in MM5 model for a cyclonic situation led to a diffusion and dilution of the cyclonic circulation in subsequent predictions (24 hour and beyond) made by the model, if the winds at upper levels did not support the evolution of a cyclonic flow. In order to achieve a coupling between the circulations at surface and other model levels, we analysed the 3-dimensional wind fields generated by the model for several cyclonic situations. A complex EOF analysis of several model generated vertical wind profiles was then carried out to determine the dominant vertical structure of wind velocity in a cyclonic situation. While assimilating the QuikScat winds for a cyclone, the above structure functions were used to determine an approximate 3-dimensional fields of wind and this complete field was then assimilated by the method of nudging. Our preliminary analysis indicates that the above scheme shows reasonably good success in retention and evolution of the cyclonic flow fields in subsequent model integration after the assimilation process.