velocity structure used is from the EMC model GLAD-M35, thus smooth at these length scales. the simulation mesh is rather coarse and so is the USGS finite-source solution. that leads to numerical noise.
discussing coda waves and source complexity can come later, try support rescue efforts now.
