Automatic arrival picking for microseismic signals based on variational mode decomposition and Akaike information criterion

Microseismic (MS) monitoring, which captures signals generated during rock mass fractures, can monitor changes in underground reservoir characteristics. It is of significant importance for the guidance and evaluation of hydraulic fracturing and prediction of geological disasters. However, the signals recorded by seismic detectors often contain various types of noise, especially in surface monitoring with more complex environments. Extracting effective MS signals and accurately picking up their arrivals serves as the foundation for subsequent positioning and other inversion processes. Given the unknown frequency distribution of effective MS signals, it is difficult to achieve signal-to-noise separation through simple filtering methods. In this paper, we propose a novel automatic arrival picking method based on variational mode decomposition (VMD) and Akaike information criterion (AIC). First, VMD is utilized to decompose the original signal into several intrinsic mode functions (IMFs). Then, the Pearson correlation coefficient (CC) and peak-to-average power ratio (PAPR) are combined to determine the effective components. Finally, we reconstruct the signal and employ the AIC method to pick up the arrival of MS events. Applying this method to synthetic tests based on Ricker wavelet, the results demonstrate that it can accurately distinguish effective signals from noise components, exhibiting superior robustness to noise compared to other arrival picking methods. Furthermore, the processing results of field MS signals during the fracturing process of a shale gas well in Sichuan Province also validate the advantages and application potential of the proposed method.