We modified the common-offset-common-reflection-surface （COCRS） method to attenuate ground roll, the coherent noise typically generated by a low-velocity, low-frequency, and high-amplitude Rayleigh wave. The COCRS o...We modified the common-offset-common-reflection-surface （COCRS） method to attenuate ground roll, the coherent noise typically generated by a low-velocity, low-frequency, and high-amplitude Rayleigh wave. The COCRS operator is based on hyperbolas, thus it fits events with hyperbolic traveltimes such as reflection events in prestack data. Conversely, ground roll is linear in the common-midpoint （CMP） and common-shot gathers and can be distinguished and attenuated by the COCRS operator. Thus, we search for the dip and curvature of the reflections in the common-shot gathers prior to the common-offset section. Because it is desirable to minimize the damage to the reflection amplitudes, we only stack the multicoverage data in the ground-roll areas. Searching the CS gathers before the CO section is another modification of the conventional COCRS stacking. We tested the proposed method using synthetic and real data sets from western Iran. The results of the ground-roll attenuation with the proposed method were compared with results of the f-k filtering and conventional COCRS stacking after f-k filtering. The results show that the proposed method attenuates the aliased and nonaliased ground roll better than the f-k filtering and conventional CRS stacking. However, the computation time was higher than other common methods such as f-k filtering.展开更多
基金the creators of the Seismic Lab, a MATLAB seismic data processing package, the NIOC Exploration Directorate, Iran for financial support and the data of the Project No. 89235
文摘We modified the common-offset-common-reflection-surface （COCRS） method to attenuate ground roll, the coherent noise typically generated by a low-velocity, low-frequency, and high-amplitude Rayleigh wave. The COCRS operator is based on hyperbolas, thus it fits events with hyperbolic traveltimes such as reflection events in prestack data. Conversely, ground roll is linear in the common-midpoint （CMP） and common-shot gathers and can be distinguished and attenuated by the COCRS operator. Thus, we search for the dip and curvature of the reflections in the common-shot gathers prior to the common-offset section. Because it is desirable to minimize the damage to the reflection amplitudes, we only stack the multicoverage data in the ground-roll areas. Searching the CS gathers before the CO section is another modification of the conventional COCRS stacking. We tested the proposed method using synthetic and real data sets from western Iran. The results of the ground-roll attenuation with the proposed method were compared with results of the f-k filtering and conventional COCRS stacking after f-k filtering. The results show that the proposed method attenuates the aliased and nonaliased ground roll better than the f-k filtering and conventional CRS stacking. However, the computation time was higher than other common methods such as f-k filtering.
