DATA

The data is from an array laid out across the Colorado-Wyoming border. The array is composed of 30 seismometers laid out in line in a semi-straight line with approximate spacing of 2.6 km. The data consists of a series of arrival times from known station locations, at known dips and arrival directions (back azimuths). The left panel of Figure  shows an initial velocity model with the stations marked with an `*'. The right panel of Figure  shows the distribution, in terms of azimuth and angle, of the earthquakes used in this experiment.

 

initial
Figure 1 The left panel shows an initial velocity model with the recording stations marked with an `*'. The right panel shows the distribution, in terms of azimuth (compared to receiver line) and dip, of the earthquakes used in this experiment. The large change at 43 km is the Moho.

A linear trend is removed from the arrival times for each event (earthquake). The subtraction of the trend is meant to account for the varying source directions (earthquake position compared to the array). The trend removed arrived times generally form the data $\bf \Delta t$ for the global tomography problem. Figure  shows these arrivals with the trend removed. Note the general shapein the times.

 

dt Figure 2 The arrival times with a linear trend removed. Note the pattern in the arrival times.