The regionalization scheme as of June 2001. Each region corresponds to a distinct geologic province within which a simple layered velocity structure is assumed.
1D velocity models which fit the empirical traveltime curves of each region.
Method
Each of the roughly 25 regions defined for Asia corresponds to a geologic province. The regions have been selected so that the
structure within any one region is relatively homogeneous. This allows a single traveltime curve to predict traveltimes throughout
the region. The traveltime curves have been been determined empirically from many decades worth of observations in many cases by
local seismic observatories. Here, we derive simple 1-D velocity profiles which predict the observed traveltimes.
Our starting model is the modified IASPEI model described in the Prelimenary IASPEI model page using the revised node gridding approach. Raytracing and traveltime prediction is performed in the full 3-D model. (Since a cartesian coordinate system is being used, a true 1-D model would not have included the effects of the spherical Earth.) The error in the predicted traveltime is inverted to obtain an improved model. A 1-D constraint is applied to the inversion by permitting each depth in the model to vary as a whole. Thus, the frechet derivatives are the partial derivatives of each traveltime, with respect to each velocity layer in the model, dT/dV. A light damping 0.1*||GTG||inf (10% of the largest value in the squared Frechet matrix) was applied to the inversion to keep it stable.
Models by region
All the traveltime curves can be fit with similar velocity models. Constraints of the raytracing require the models to be defined
as linear gradients between irregularly-spaced depths. The crust is represented as a single layer with velocity increasing from
4-5.5 km/s at the surface to 7.5-8.5 km/s at the Moho. The models include the transitions at 210 km, 410, and 660 km though only
the 210 has a significant impact on regional-distance seismic waves.
The regionalization scheme as of June 2001. Each region corresponds to a distinct geologic province within which a simple layered velocity structure is assumed. |
1D velocity models which fit the empirical traveltime curves of each region. |
Traveltimes and errors
The plot below on the right shows the traveltimes and prediction errors for each of the regions. In most cases the 1-D structures
predict first arrival traveltimes to better than 0.2 seconds and often better than 0.1 seconds.
 The figure at right shows the velocity profile for each region compared to the average velocity structure (all 22 regions averaged together, NOT weighted by area.) The average structure is shown at the top of the figure. The odd spikes in the velocity difference plots is caused by the varying depth of the crust-mantle interface. |
 Traveltimes for each region. The observed traveltime is the empirically-derived curve of Khalturin et al. This is accompanied by the traveltime predicted from the1-D structures shown above. An IASPEI curve is shown for reference. |
Data: traveltimes and models for each region
Region perimeter: Perimeter of region as defined by closed polygon lat and lon coordinates
traveltime data: Data as provided by MRC in June 2001. Files are not re-interpolated though a 0 s traveltime at range 0 has been added.
1-D velocity model: Best fitting (spherical Earth) 1-D velocity model for the region.