Source Location And Source Mechanism Inversion Of Surface Microseismic Monitoring

At present,microseismic monitoring technology plays an important role in the exploration and development of unconventional oil and gas reservoirs,which makes microseismic monitoring technology become a popular geophysical technology with a broader development prospect.In recent years,surface microseismic monitoring has received more and more attention.Surface microseismic monitoring has the advantages of low acquisition cost,more flexible geophone embedding,wider azimuth coverage and so on.Coupled with the improvement of geophone performance and the development of surface microseismic technology,the surface microseismic monitoring technology has gained more and more recognition.This paper mainly studies several important steps in the processing of surface microseismic data,such as noise suppression of surface microseismic data,automatic picking of microseismic events,automatic localization of surface microseismic events and an exploratory study on source mechanism inversion of surface microseismic monitoring is presented.One of the main characteristics of surface microseismic monitoring data is low signal-to-noise ratio,weak energy of effective signals,strong energy of noises and rich types of noises.Therefore,noise suppression is an indispensable step in the processing of surface microseismic data.Conventional noise suppression methods use the difference between signal and noise in the transform domain,such as time domain,space domain,frequency domain,time-frequency domain and frequency-space domain and so on,to suppress noise.Based on the theory of multi-scale morphology,this paper develops an improved multi-scale morphological denoising method.Conventional multi-scale morphology requires artificially identifying the distribution of effective signals on multi-scale profiles in the process of suppressing noise,while the noise in surface microseismic data is complex and diverse,and the distribution of effective signals and noise on multi-scale profiles is also difficult to find out the specific rules.The STA/LTA method is the most commonly used method for microseismic signal recognition,and is widely used due to its advantages such as easy implementation and high recognition efficiency.In this paper,the STA/LTA method is combined with multi-scale morphology to solve the difficulty of artificially identifying effective signals on multi-scale profiles,and a fully automatic multi-scale morphological denoising process is realized.For surface microseismic data,although the signal-to-noise ratio can be improved to a large extent through noise suppression,the energy of microseismic signals transmitted to the surface that can be collected by geophone is very weak and the number is very small because the microseismic signals themselves are very weak.For a large number of continuous surface microseismic records,it has always been a difficult problem to identify effective signals automatically and quickly and accurately.Based on the theory of multi-scale morphology,this paper develops a feature function method based on multi-scale morphology.This method,combined with cross-correlation technique,can accurately identify the components of effective signals on multi-scale profiles,so as to realize the automatic recognition of surface microseismic signals.In order to realize the purpose of microseismic monitoring,the identified microseismic events need to be located accurately.Therefore,surface microseismic location is the most critical part of surface microseismic data processing.Due to the low signal-to-noise ratio of surface microseismic data,it is difficult to obtain accurate velocity model,which leads to the low accuracy of surface microseismic location.In this paper,a series of localization methods based on local equivalent path are studied,including basic local equivalent path method,iterative local equivalent path method and calibrated iterative local equivalent path method.The basic local equivalent path method is based on the theory that the propagation path of two adjacent events to the same receiving point has a high similarity near the surface.In this way,the path of the master event can be used to replace the path of the target event to improve the positioning accuracy.The iterative local equivalent path method can effectively eliminate the positioning error caused by the replacement path.The calibrated iterative local equivalent path method can greatly reduce the dependence of positioning results on the velocity model and has the advantages of high degree of automation,good noise robustness,and short time consumption.The localization method proposed in this paper can be extended to such fields of locating based on signal travel as natural earthquake,volcanic earthquake and so on.On the basis of deducing the loading mode of moment tensor source in finite difference scheme,the concept of wave field function of unit couple is proposed and explained,and then the inversion method of seismic source mechanism based on wave equation is improved by using this function.