摘要
以模型试算的结果为依据,对薄层频谱的理论计算值、薄层直观的分辨率与信噪比、时间-振幅解释量板、截断效应与时窗选取、穿层对时频特性的影响、地层横向变化对时频特性的影响、薄层的自相关和自回归特征等问题进行了定量研究和重新认识。研究发现:①理论上,随着薄层的厚度变薄,薄层的能量变小,频率有所提高;②对于常规地震资料,薄层分辨率在10 m数量级,开展薄层研究,信噪比最好要大于2;③通过对调谐厚度与速度和子波频率关系的研究,发现实际调谐厚度要小于1/4波长,振幅解释量板的可靠性要好于时间解释量板;④对常用的5种窗函数的分析表明,布莱克曼窗最好;⑤对于穿层严重的薄储集体,波形特征往往会受上覆和下伏地层影响,但其频谱特征基本不受影响;⑥对于地层横向变化较大的地层,时窗开取的方式不同,其频谱横向变化也不同,一般以上下开时窗效果较好;⑦通过薄层自相关和自回归特征研究,发现薄层特征呈比较好的规律性。
According to the calculating results of the models, this article discusses some thin-bedding questions, such as the theoretical calculating value of frequency spectrum, the instinct resolution and the signal/noise ratio, truncation effects so on, the effects in the time-frequency characteristics for the vertical or lateral variations of layers, the thin-bedding characters gotten by autocorrelation and auto-recursion. We have obtained some conclusions: 1) When the thickness of the thin-beddings becomes fatter, its energy will be smaller and its frequency will be higher. 2) For normal seismic data, the thin-bedding resolution is 10m, the signal/noise ratio would better be bigger than 2. 3) About the studying of the relationship between tuning thickness and velocity or wavelet frequency, we can find that fact the practical tuning thickness is smaller than the theoretical one and the reliability of amplitude interpretation template is superior to time interpretation template. 4) The Blackman's is best among the five window's functions. 5) For the thin reservoir in which vertical variations are serious, the waveform characteristic always are affected by the layers up and down, but the characteristic of its frequency spectrum are not affected. 6) For the layers in which lateral variations are serious, lateral variations of frequency spectrum vary with the time window's choosing. 7) By the studying of autocorrelation and auto-regression, we find the orderliness of the thin-bedding characteristics is very good on some certain extent.
出处
《石油物探》
EI
CSCD
2004年第6期541-546,共6页
Geophysical Prospecting For Petroleum
关键词
薄层
地球物理特征
地层
横向
地震资料
子波
储集体
实际
理论计算值
发现
<Keyword>in bed
windows function
resolution
signal/noise ratio
interpretation template
time-frequency characteristics
autocorrelation
auto-regression