P-velocity, radially aniostropic, 3-dimensional mantle model 8,
after Boschi & Dziewonski (JGR, 2000). This model is obtained from
P, PcP, PKP data (based on ISC bullettin after appropriate corrections),
with the assumption that the outer core be spherically symmetric.
Of all the models published by Boschi & Dziewonski (JGR, 2000), we
consider this one to be the most realistic.

Approximately equal area block model (5 degrees X 5 degrees X 193.3 km).
It includes two images of the mantle, one for horizontal P-velocity
(files h*.d) the other for vertical P-velocity (files v*.d);
it also includes an image of the core-mantle boundary topography
(file cmb.d).

The mantle files are defined as for model BDP98:
the layers are given in the files vlaye_01.d through vlaye_15.d, 
and hlaye_01.d through hlaye.15.d in the following format (example):

      1.25403E-02
   0.    85.0000
   0.    90.0000
     90.0000    90.0000
     90.0000    85.0000
   -0.301233
     90.0000    85.0000
     90.0000    90.0000
     180.000    90.0000
     180.000    85.0000
   -0.391459
     180.000    85.0000
     180.000    90.0000
     270.000    90.0000
     270.000    85.0000

the single numbers are the value of the PERCENT velocity perturbation
relative to PREM within each block; the following four pairs of numbers
are longitude and latitude of the four vertices of the corresponding
block... and so on...

I believe that the best way to plot this file is to use gmt.
First you need to write a program to turn each file into something like:

> -G165/243/172
   0.    85.0000
   0.    90.0000
     90.0000    90.0000
     90.0000    85.0000
> -G255/184/101
     90.0000    85.0000
     90.0000    90.0000
     180.000    90.0000
     180.000    85.0000
> -G255/162/071
     180.000    85.0000
     180.000    90.0000
     270.000    90.0000
     270.000    85.0000

where the three integers after -G, separated by /, define the color
corresponding to the velocity perturbation in the block (and then
the block-vertices coordinates are just the same), based on a colorscale
of your choice.
Assuming that you called this new file image, you can use (more or less)
the commands

psxy image -JH180/6 -Bg45/g45 -R0/360/-90/90 -K -M >! fig.ps 
pscoast -R0/360/-90/90 -O -JH180 -W1 -Dc -A5000 >> velocity.ps

to obtain a figure with that layer of the model (and the coastline).

For comparison, the postcripts files vh.ps (horizontal P-velocity)
and vv.ps (vertical P-velocity) are given, which include the lateral 
velocity perturbations for all the layers in our model, in order
of increasing depth (1--15) from left to right and top to bottom.
The color scale saturates at 1%. 

The CMB file cmb.d is given in the same format and can be plotted in
the same fashion. The CMB topography anomalies are defined as elevation
with respect to the constant CMB radius of PREM; in practice, a positive
topography anomaly corresponds to elevated CMB, while a negative anomaly
corresponds to depressed CMB.

Compare with the figure cmb.ps (where the average cmb perturbation
was set to zero); the color scale, there, saturates at 10km.


Lapo Boschi, Cambridge, July 25, 1999.

boschi@seismology.harvard.edu