Synthetic microseismic datasets

Joe Wong, Peter Malcolm Manning

We designed and created synthetic microseismic datasets for testing the accuracy of different source location algorithms. Three-component microseismograms with sources and receivers embedded in horizontally layered velocity structures were generated either by ray-tracing (RT) and convolution, or by 2D finite-difference (FD) modeling. The RTmodeled seismograms include direct and head-wave arrivals, but no reflected arrivals. The ray-tracing method can approximate some effects of TTI anisotropy in the layers. Traces from FD modeling include reflected as well as direct and head wave arrivals, but they do not include effects due to anisotropy. Gaussian noise and harmonic noise are added to the synthetic microseismograms to simulate different signal-to-noise levels, and the modeled data are stored in SEG2 format to conform to field-recorded data files.