To improve the utilization of infrared satellite data, more data should be assimilated into the initial field of numerical forecasting mode so as to minimize the influence of cloud pollution. According to the data from Atmospheric Infrared Sounder (AIRS), a cloud detection method is used to identify Genting Height. After the validity of the method is verified, all of the channels above cloud top are retained in the cloud region of a large-scale satellite field of view and the brightness temperature data of these channels are assimilated into the initial field of the numerical prediction mode. By comparing and analyzing the simulated paths of the typhoon by different schemes, the influence of the cloud detection method on Typhoon numerical simulation is evaluated. The results show that the detection method is better in detection. The distribution of the clear sky and cloud area detected corresponds to the actual observation and the high cloud top area detected corresponds well to the typhoon cloud system. After the channel brightness temperature above cloud top is utilized, the utilization of infrared satellite data is improved greatly. The data utilized is 2.4 times as those of the original data. In typhoon path simulation, the increase in the amount of data makes the path simulation more accurate and is obviously close to the observed values after assimilation of above channel data.