Inverting both group and phase velocities results in models sharing characteristics of each of the separate inversions - the low velocity zones at 25-35 km depth and the low velocities in the upper mantle. However, the low velocity upper mantle is suspect. Evidence suggest a velocity gradient in the upper mantle. Since this inversion has constrained the mantle to be only two-layered, the resulting models may reflect an attempt to accomodate this gradient with only two layers. Inversions based on model 2 (shown elsewhere) allow more degrees of variance in the mantle.
The group and phase velocities are fit reasonably well in this inversion. The predicted phase velocities collapse to almost the same line which is a smoothed version of the real data. This suggests that the "bumps" in the measured phase velocity curves are measurement error and do not reflect structure.
It is remarkable that the final models predict almost identical velocities despite the range of model values.
| MODEL PLOT: range of initial models: gray final models: violet mean of final models: black VELOCITY PLOT: Initial group velocity: pale blue Initial phase velocity: pink/gray Final models: black Measured group velocity: blue Measured phase velocity: red |
