摘要
尽管人们采用了许多不同的反射系数模型来更好地模拟地下介质物性参数的变化情况,目前尚没有一种理想模型能够真实地反映反射系数的复杂特征,这是因为反射系数本身在统计上具非高斯性,功率谱为有色谱,同时为多重分形特征。根据从波阻抗测井资料得到的真实反射系数的一般特征来讨论反射系数模型的适用性。结果表明,反射率的多重分形特性使得很多看似合理的模型具有缺陷。多尺度反褶积方案可用来避开传统预测反褶积中所要求的一些假设,小波基础上的多尺度反褶积研究具有理论和实际上的重要意义。
Although various reflectivity models have been proposed to better simulate true subsurface property variations, none of them are ideal comparing to the complexity of reflection sequences, which are non-Gaussian, with blue spectrum and multifractal. In this paper, applicable reflectivity models are assessed against the general properties of true reflection coefficients derived from a coustic impedance logs. It is found that the multifractal requirement of reflectivity sets many appealing models flawed. To circumvent these deficiencies and difficulties, multiscale deconvolution schemes are preferred, which may relax many assumptions traditionally required in predictive deconvolution.
出处
《勘探地球物理进展》
2005年第4期250-257,共8页
Progress in Exploration Geophysics
基金
美国勘探地球物理学会奖学金及同济大学211工程人才引进启动基金的资助。
关键词
反褶积
反射系数模型
多尺度反射系数
小波变换
分形
deconvolution
reflectivity model
multiscale reflectivity
wavelet transform
fractal