A numerical study of the interaction between two simultaneous storms: Goni and Morakot in September 2009

Significant anomalous tracks were observed when the severe tropical storm Goni(0907) and typhoon Morakot(0908) in September 2009 were evaluated in short distances.The relationship between the two is regarded as a case of binary interaction.Based on an MM5 model(fifth-generation mesoscale model of Pennsylvania State University-National Center for Atmospheric Research),in this study a series of sensitivity experiments were designed to determine the binary interaction between them.The sensitivity of the storm characteristics to the binary interaction was demonstrated through modeling experiments with different TC intensities and sizes based on the bogus vortices initialization.Furthermore,the contributions of large-scale environmental flow and the effects of interaction between the motions of the cyclones were distinguished by numerical experiments using only one of the TC vortices.Results from these experiments show that Morakot(0908) had a greater impact on the motion of Goni(0907),whereas Goni(0907) had a relatively limited impact on Morakot(0908).At the upper level,the northeasterly jet flow in the third quadrant of Morakot(0908) enhanced the upper-level divergence of Goni(0907) and had an important role in maintaining and increasing Goni’s(0907) intensity.And at the lower level,Morakot(0908),with strong convergence and ascending airflow,made a stable transport channel of southwesterly warm and wet flow,thus supporting the lower-level water vapor convergence of Goni(0907).Goni(0907),which was located upriver of the southwesterly flow,intercepted part of the water vapor transportation in the southwesterly flow,causing the water vapor convergence to strengthen while the water vapor convergence of Morakot(0908) weakened.     

基于支持向量机的CloudSat卫星云分类算法

从基于云角色的分类思想出发,利用星载毫米波雷达探测资料提取云的特征参数,建立支持向量机（support vector machine,SVM）模型实现云的分类。通过与BP（back propagation）网络模型的分类结果进行对比,发现两种模型都具有较好的分类能力,但SVM模型的识别准确率更高,计算速度更快。基于CloudSat资料的云分类实例表明,SVM模型的分类结果与CloudSat数据处理中心（Data Processing Center,DPC）发布产品具有很好的一致性。     

基于神经网络的卫星高度计海面风速反演算法

通过对TOPEX/Poseidon高度计资料与NDBC浮标实测数据进行时空匹配处理,得到同步数据集,利用人工神经网络方法试验得到海面风速反演算法,并与业务运行的M CW算法进行分析比较,指出考虑波浪状态影响因素的神经网络算法在均方根误差和对称性方面的优越性。研究表明利用神经网络方法反演海面风速是可行的。     

云相态的卫星遥感研究进展

云相态不仅是气象学和气候学研究的重要参量,也是卫星云参数反演的关键要素,其识别的准确性直接关系到光学厚度、有效粒子半径等光学和微物理参数的反演精度.卫星遥感技术的发展为云相态的识别提供了前所未有的技术手段.对能够提取云相态信息的主要星载遥感源进行了介绍,重点总结了近30年来国内外利用星载被动光学遥感资料、偏振辐射资料以及主动雷达资料进行云相态反演的基本原理、方法以及技术特点,最后针对目前云相态反演工作中存在的一些问题和发展趋势进行了探讨.     

西北太平洋双热带气旋相互作用统计分类及其特征分析

采用中国气象局与上海台风研究所提供的1949-2007年西北太平洋热带气旋最佳路径资料,统计分析了西北太平洋上的双(多)热带气旋相互作用现象.针对统计出的双热带气旋样本,根据相互作用两气旋的单独移动路径和它们相互作用的路径特征,归纳得到7种相互作用类型( A～G),并用聚类分析方法拟合得到每种类型的平均回归路径.分析讨...