上海浦东地区“梅雨期”降水及其多尺度时频特征

提出了浦东地区“梅雨期”降水的概念。在此基础上,利用小波变换分析了浦东地区“梅雨期”降水的多尺度时频特征及其突变现象。结果表明:浦东地区6月1日—7月10日期间的降水序列能更加客观地反映该地区春末夏初这一特殊的“梅雨期”降水量的实际状况;浦东地区“梅雨期”降水存在准2 a和10～20 a的主要振荡周期;利用小波逆变换重构的不同振荡周期的时间序列能更好地反映原序列的主要周期振荡特征及其趋势信息。     

Simulation of nose whistlers: An application to low latitude whistlers

Simulation technique for whistler mode signal propagating through inhomogeneous plasma using WKB approximation has been developed (Singh, K., Singh, R.P., Ferencz, O.E., 2004. Simulation of whistler mode propagation for low latitude stations. Earth Planet Space 56, 979-987). In the present paper, we have used it for the analysis of recorded signals at low latitudes and estimated the nose frequency, which is not observed on the dynamic spectra. At low latitudes nose frequency is ∼100 kHz or more and therefore it is absent in the dynamic spectra due to attenuation of the signal at higher frequencies. The importance of nose frequency is in determining the exact path of propagation, which is required in probing of ambient medium. It is shown that the method permits to study the nose frequency variation, it can be used to deduce physical parameters as the global electric field. A case study permits to get a reasonable value of the electric field, which up to now could not be done at very low latitude.     

危险率函数在地表破裂型地震危险性评估中的应用—以宁夏灵武断裂为例

在对中国北部板内地区的古地震资料拟合、给出时间相依的统计模型基础上,利用近期获取的灵武断裂有关资料和危险率函数,对灵武断裂未来一定时段的发震潜势和当前地震危险性特征进行了分析,认为该断裂已进入发震时段,在未来10－100年内有发生7．0－7．5级左右地震的潜势。     

基于小波变换的多光谱图像压缩方法

在分析多光谱图像小波变换后系数特点的基础上,提出了一种共享有效图的小波变换压缩方法（ＳＳＭＷＴ）。该方法将小波变换压缩技术中的零树编码推广到多光谱图像压缩中,利用多光谱图像的结构相关性,对多幅小波图像只需构造一幅有效图,同时去除空间冗余和谱间结构冗余,并与Ｋ－Ｌ变换相结合,进一步去除谱间统计冗余,实验表明了该方法的有效性。     

Localized multiscale energy and vorticity analysis: I. Fundamentals

A new methodology, multiscale energy and vorticity analysis (MS-EVA), is developed to investigate the inference of fundamental processes from oceanic or atmospheric data for complex dynamics which are nonlinear, time and space intermittent, and involve multiscale interactions. Based on a localized orthogonal complementary subspace decomposition through the multiscale window transform (MWT), MS-EVA is real problem-oriented and objective in nature. The development begins with an introduction of the concepts of scale and scale window and the decomposition of variables on scale windows. We then derive the evolution equations for multiscale kinetic and available potential energies and enstrophy. The phase oscillation reflected on the horizontal maps from Galilean transformation is removed with a 2D large-scale window synthesis. The resulting energetic terms are analyzed and interpreted. These terms, after being carefully classified, provide four types of processes: transport, transfer, conversion, and dissipation/diffusion. The key to this classification is the transfer–transport separation, which is made possible by looking for a special type of transfer, the so-called perfect transfer. The intricate energy source information involved in perfect transfers is differentiated through an interaction analysis.The transfer, transport, and conversion processes form the basis of dynamical interpretation for GFD problems. They redistribute energy in the phase space, physical space, and space of energy types. These processes are all referred to in a context local in space and time, and therefore can be easily applied to real ocean problems. When the dynamics of interest is on a global or duration scale, MS-EVA is reduced to a classical Reynolds-type energetics formalism.     

基于Db小波与中值滤波的安检图像噪声消除算法

设置成像安检系统是机场等重要场所通常采用的安全措施。根据CT安检图像既含高斯白噪声又含脉冲噪声的特点，运用将Db小波变换与中值滤波相结合的综合去噪方法，以同时滤除高斯噪声和脉冲噪声，以获得更好的视觉效果。     

希尔伯特-黄变换中拟合过冲和端点飞翼的原因及解决办法s

深入分析了现有的三次样条拟合方法，指出ldquo;拟合过冲rdquo;现象产生的原因，并提出新的三次样条拟合方法和正弦函数拟合方法. 这两种新拟合方法都具有消除ldquo;过冲rdquo;现象，拟合函数连续、一阶导数连续、计算简便、计算量小、易于操作等优点. 同时，指明了ldquo;端点飞翼rdquo;现象产生的原因，并提出一种处理资料曲线的端点问题的技巧. 该技巧能够在不改变资料曲线的条件下消除ldquo;端点飞翼rdquo;现象. 拟合新方法和新技巧能结合运用，得到完全没有ldquo;拟合过冲rdquo;和ldquo;端点飞翼rdquo;的拟合曲线，为时序曲线固有模态函数分解提供了可靠的技术基础.      

图像的小波数字水印添加与检测

提出了一种基于小波变换的图像加水印及检测的新方法,即对图像文件进行二维多尺度离散小波变换,得到图像数据高频和低频部分。在高频部分数据中,加入随机数水印并进行图像重构,检测时以同样方法进行小波分解,检测高频部分与随机数水印的相关性。此方法可用于图像防伪。     

拟神经网络反演在煤田中应用

Parker快速富氏变换反演单一密度界面存在两方面问题：一是求解目标界面重力异常难度较大;二是受正演速度及反演参量维数的限制,不能对界面进行精细划分.拟神经网络BP算法的引入,首先解决了快速三维正演问题,又突破了反演参量维数的限制,实现快速收敛,有效解决两个或多个密度界面的反演问题.在实际应用中,先用密度“补偿法”正演求取剩余生力异常,然后利用拟神经网络BP算法同时反演两个二维密度界面,拟合求得两个界面的深度异常,在此基础上预测煤田.