The first seismic arrivals from a distance earthquake are understood to follow a Great Circle path. The Great Circle is the circumference on the Earth’s surface that contains both the earthquake epicenter and the seismograph location. By definition, the center of the Great Circle is the center of the Earth. There are two Great Circle paths from the epicenter to the seismograph, one going each direction along the circle. The short one (minor arc) is usually the one of interest.
Following the Great Sumatra-Andaman Island Earthquake in 2004, I remember seeing a plot (called a seismic record section as described here) of seismic waveforms from many world-wide stations which clearly showed the propagation of the surface waves as a function of time and distance. This plot also clearly showed the arrival of the surface waves along the other Great Circle Path, the long path. The IRIS website that I referenced in the last post has this type of plot for the the M8.9 Honshu earthquake.
I decided to look for this long path arrival in the waveform data recorded from my Lehman seismograph. Our short path distance is shown in the arrival time calculations listed (as delta) at the end of my first post on this earthquake as 91.68 degrees. I extracted an approximately three hour segment of the waveform and used the filtering features of AmaSeis to clean up the signal. I used the long period filter and entered in parameters that I thought might be appropriate for my seismograph, although I admit to not fully understanding this particular filter. Of particular interest was the time of the first surface arrivals which I picked at 6:24:44 or 2301 seconds after the earthquake struck. This is a little (about 25 seconds) sooner than the arrival time calculator had given. Then I calculated the velocity along the short path to be 0.0398 degrees per second. The long path distance is just 360° minus the short path distance or 268.32°. Dividing the long path distance by the velocity along the short path (which is the same along the long path) indicated that the long path arrival should show at 6734.34 seconds or 112.239 minutes after the earthquake struck Japan. My best pick of the arrival was at 113.03 minutes so I’m pretty certain that is what I am looking at.
On my seismograph, the long path arrival will probably only show up on the most powerful earthquakes such as this one.