Detection of karst collapses through microtremor surface waves based on windowing cross-correlation function
The study area is located in the karst development area of Pingxiang in the west of Jiangxi Province,China.The landform of this area is complex,low in the northwest and high in the southeast.The fold action and gravitational sliding action led to the development of faults and extensional sliding nappe structures in the area,accompanied by magmatic intrusion activities,which has formed multi-phase superimposed complex structures.Atmospheric precipitation and groundwater in the upstream limestone areas constitute the main water source in the study area.Meanwhile,karst collapses may cause the formation of holes below the surface,seriously endangering people's life and property.Therefore,finding out the geological situation of the collapse area can provide a reference for the understanding of the geological characteristics and the groundwater system in this area.Microtremor is a kind of persistent weak vibration signal observed on the surface caused by industrial vibrations,traffic noises,tidal currents,atmospheric activities and other activities on the earth.Surface waves are formed by vertical waves and transverse waves interfering on the surface.They have the characteristics of low speed,low frequency and frequency dispersion,which lay a foundation for the detection of underground structure.The method of microtremor surface waves survey utilizes various types of vibrations that continuously exist in nature as signal sources.It extracts the information on seismic surface wave field from microtremor records and uses this type of information for imaging underground media.The steps include microtremor signal acquisition in the study area,data preprocessing,empirical Green function calculation,extraction of surface wave dispersion curves and inversion of velocity structure for transverse waves.Among these steps,the empirical Green function is obtained through the cross-correlation operation of the microtremor signals recorded by two detectors,and calculating the empirical Green function is the key to obtain the surface wave information.This study detects the karst collapses in the study area by using microtremor surface waves.However,the surface wave signals are affected by uneven distribution of natural noise sources and random noises,which may cause the low signal-to-noise ratio of the Green function.Therefore,the direct use of the empirical Green function for subsequent data processing may get the dispersion energy spectrum with low resolution,which is not conducive to the subsequent extraction of high-quality dispersion curves and accurate inversion.Due to the above shortcomings,this study first optimized the window function for the cross-correlation function of microtremor signals to improve the signal-to-noise rate of microtremor data,and to enhance the resolution of energy spectrum of microtremor surface wave dispersion.Then,the extraction of dispersion curves and inversion of the virtual source surface record of each group in the study area were conducted to obtain the transverse wave velocity structure of each point along the measurement line.Finally,the distribution positions and depths of karst collapses in the study area were revealed,according to the the transverse wave velocity section below the measurement line generated by inversion as well as the geological interpretation of drilling data.The results show as follows,(1)The selection of window function in the cross-correlation function calculation will affect the resolution of the dispersion energy spectrum,and the window function should be tested in processing the microtremor data.(2)The processing and optimizing of window function for the cross-correlation function can effectively improve the signal-to-noise rate of microtremor surface wave and the resolution of the dispersion energy spectrum,widen the frequency band range of dispersion curves,and improve the accuracy of the inversion.(3)The method of microtremor surface waves survey is highly applicable to karst collapse detection,and the optimization of the window function can determine the range of underground hazards in a more accurate way.
万方数据
维普
