>Try changing around the order of the deltas.
>One of the 6 combinations should work for you.
Although I couldn't see why this should help I tried it out, and it did
not work. As I understand it,
# delta x points in (x,0,0) direction in an (x,y,z) system
delta 25.4237 0 0
# delta y points in (0,y,0) direction in an (x,y,z) system
delta 0 26.9231 0
# delta z points in (0,0,z) direction in an (x,y,z) system
delta 0 0 -6.6667
If I have a 3D-vector dataset where the points are stored in an (x,y,z)
grid as indicated by
write (30) (((vx(i,j,k),vy(i,j,k),vz(i,j,k),
& k=1,nz),j=1,ny),i=1,nx)
(and not in a, say, y,z,x one), and where the components are also stored
by this convention, I must expect the program to reproduce this
coordinate ordering when I specify the grid format in the header as
above; everything else would be completely arbitrary and make no sense
at all.
I have added axis labels to the snapshots; in detail:
http://www.geophysik.uni-frankfurt.de/~ruedas/correct.jpg
shows the desired output, produced with the screwed write ordering
write (30) (((vy(i,j,k),vz(i,j,k),vx(i,j,k),
& k=1,nz),j=1,ny),i=1,nx)
(note the shifted components!).
http://www.geophysik.uni-frankfurt.de/~ruedas/false.jpg
shows the undesired output, produced by the right write ordering
write (30) (((vx(i,j,k),vy(i,j,k),vz(i,j,k),
& k=1,nz),j=1,ny),i=1,nx)
--
------------------------------------------------------------------------
Thomas Ruedas
Institute of Meteorology and Geophysics, J.W.Goethe University Frankfurt
e-mail: [EMAIL PROTECTED]
http://www.geophysik.uni-frankfurt.de/~ruedas/
------------------------------------------------------------------------
