地震信号的吸收滤波与储集层含气性识别

地震波在烃类储集层中传播时，反射频率的吸收衰减响应在原始地震记录中不容易被发现。基于ARMA（Autoregressive moving average)模型时间序列分析方法和Prony 变换理论的吸收滤波技术，采用阻尼谐波对地震信号进行分解，能够产生振幅、频率、相位和吸收系数等参数的不同分量，这些分量构成了子波信号库。结合测井、岩石物理、地质等资料，对库存子波信号进行综合分析、筛选、重构，从而生成优化重组了的地震信号。与原始地震信号相比，吸收滤波数据能够更有效地刻画地质体的各种地震响应，有利于更直观、更高效地实现烃类储集层的预测。将该技术应用到理论模型试算和川西坳陷深层须家河组气藏的含气性识别中，均获得了良好的应用效果。

Application of Absorption Filter on Seismic Signals to Gas Reservoir Identification

When seismic wave propagates in hydrocarbon-bearing reservoirs, its reflecting frequency in response of absorption and attenuation is difficult to be detected in original seismic data. The absorption filter technology of seismic signals is based on autoregressive moving average (ARMA) analysis method and Prony transforming theory to decompose seismic signals into damping waves which are different in amplitude, frequency, phase and absorption coefficients. These decomposed signals comprise of the signal storage. We can analyze, select, reconstruct and optimize the seismic signals, combined with the information of well logging, petrophysics and geology, etc. Comparing with original seismic signals, the absorption-filtered data can depict various seismic responses more effectively, thus predicting hydrocarbon-bearing reservoirs more intuitionisticly, more effectively than using the original signals. The case study indicates that good applied result has been obtained using this method in test of model data and in gas reservoir identification.