Researches prove that TC (tropical cyclone)intensity is mainly determined by the top cloud convection strength,distribution and size.Then how to extract this information from TC cloud image is very impor-tant for TC intensity estimation.In 1988,Adler put forward a method named CST (convective-stratiform technique)to extract tropical convective cores from TC cloud image.Using MTSAT (multi-functional transport satellite)IR1 black body temperature data,the TC cloud top strong convection is extracted. Based on the convective cores number,convective core distance to TC center and convective core black body temperature extreme value,which are closely related to TC intensity,a TC intensity (expressed by V max ,the maximum sustained wind speed near surface TC center)estimation model is put forward using stepwise regress method.The experiment result shows that there is a linear correlation between their esti-mation error and their intensity forV max>40 m·s-1 andV max<18 m·s-1 samples.So according to the es-timation error distribution a linear revision is carried out. Statistical tests show this model is equivalent to Dvorak method and AMSU in TC intensity estimation accuracy.It’s also reliable based on the relationship between the convective cores,convective cores distri-bution,brightness temperature and TC intensity.It could be used in all TC life span automatically and ob-jectively,so it could be applied in the operation. Comparing with the advanced objective dvorak technique (AODT)and objective Dvorak technique (ODT),this algorithm gives accurate results in the Western North Pacific,but it’s simpler with no com-plicated pattern types identifying process or other rules.A fixed radius of 135 km area is used as TC con-vective cores searching effective area in the model,but actually the maximum wind speed radius of the TC is variable,the TC size and the strongest convective area size are different for different TC in different stage.So using the fixed searching area may affect TC intensity estimation accuracy.The research on how to get the dynamical maximum wind speed radius would be carried out in the future.%利用日本 MTSAT（multi-functional transport satellite）红外亮温资料，提取热带气旋云团中云顶较高、对流较旺盛的深对流信息，根据提取的对流核数量、对流核距热带气旋中心距离、对流核亮温极值等信息表征热带气旋强弱，初步建立了热带气旋强度估测模型；并根据该估算模型的误差分布对强度（用最大风速表示）大于40 m·s－1和小于18 m·s－1的样本结果进行了线性修正，修正后的结果与中国气象局《热带气旋年鉴》热带气旋最佳路径资料比较得到非独立样本和独立样本的强度平均绝对误差分别为5．5 m·s－1和5．9 m·s－1，均方根误差分别为6．9 m· s－1和7．7 m·s－1；对于热带低压、强台风及以上的估计平均绝对误差分别降至4．9，4．7 m·s－1，准确度较好。试验表明：利用热带气旋云团中的对流核数量、分布、冷暖与其强度建立的统计关系模型是可行的，该算法的估算精度与 Dvorak 方法、AMSU（advanced microwave sounding unit）定强算法相当。
展开▼
