Re: [ferret_users] walker circulation

[Ryo Furue <furue@xxxxxxxxxx>]

Hi Saurabh and  Ektaraful and Ferret users,

> vector . . .  uu1[y=-10:10@ave], -100*ww1[y=-10:10@ave]

I've been bothered by this kind of factor.  In this case, it's "100".

You think vertical velocity is too small and so you want to magnify
it.  But what factor is optimal?

I think that's the wrong question.  In the x-z plane, the arrow has to
be parallel to the local streamline (the local trajectory of the fluid
parcel).  Assuming that you use consistent units between the velocity
components and the axes, Ferret gives you the "correct" plot with

   vector/ASPECT  u,w

If the horizontal axis is kilometers or degrees and u is in m/s, for
example, you have to scale u appropriately (not tested).

The same principle should apply to pressure coordinates.  If the
"vertical" velocity is in the units of pressure/time, ASPECT will give
you "correct" vectors in that the fluid parcel moves parallel to the
arrows in the x-p space.  Likewise, if the vertical axis is log(p),
then "velocity" log(p)/time should be the appropriate "vertical"
velocity.  But I haven't tested these latter ideas.

Regards,

Ryo