Some more info on GMT's grdfft -T option from the GMT-Help archives :

Quoting from the GMT mailing list: > Dear GMT users,
> I am looking for more information about the grdfft -T option.
> 1) How does it decide where to use the water and infill densities?

From grdfft.c:
They get used in differences.
infill (ri) water (rw) load (rl) mantle (rm)
rigidity_d = (youngs_modulus *te*te*te)/(12.0*(1.0 - poissons_ratio*poissons_ratio))
d_over_restoring_force = rigidity_d/((rm -ri)*normal_gravity)
airy_ratio = -(rl -rw)/(rm -ri)

> 2) The example in the Reference manual pages uses -C4000. Does this mean 4km
> beneath the surface? This seems like an odd exercise, does anyone know a
> reason for this?
-C Upward (for zlevel > 0) or downward (for zlevel < 0)
continue the field zlevel meters.
The -C option is for upward or downard continuation of the field. -C4000
upward continues the field 4000 meters, which is about the depth of the ocean
in the example - so the output field should be as if it was at the surface of
the ocean.
> 3) How does it compare to other programs such as AIRYROOT?
No idea. Don't have "airyroot".
> 4) Any suggestions for references that explain this a bit further?
Geodynamics: Applications of continuum physics to geological problems
by Turcotte & Schubert, 1982, John Wiley, New York
(textbook for course in tectonophysics that I took a few years ago)

Ben Horner-Johnson

ps2pdf (Ghostscript) turns out to be buggy in some special cases where a landscape oriented GMT plot is converted to PDF. It just ignores the landscape orientation and makes a portrait one of it. This might be due to a bug in (an older version of) Ghostscript. Anyway, Apple's native version of ps2pdf "pstopdf" takes slightly longer in processing, but does the job correctly.

In order to get an inverted axis for xy plots use a minus sign in the -J option (scale/length of axis), e.g. -Jx-0.5/-1