ncl routines for divergence of moisture

Dear NCL users,

I would like to calculate the vertical-integrated divergence of
moisture div(qU), as well as its components q.div(U) + U.grad(q).
I am using the UK Met Office HadGEM GCM, which has a regular lat/lon
grid. I saw that ncl proposes few built-in functions that calculates
the divergence on Gaussian and Fixed grids using spherical harmonics.
I tried those, just to see what happens, but because my grid is
neither gaussian nor fixed, the output field is very noisy. I found
another routine, which does the calculation for any grid, uv2dv_cfd.
The output field seems right and smooth, but this function uses
centered finite differences, which is less accurate than with
spherical harmonics, so maybe this is not ideal yet.

Moreover, to calculate then grad(q), the only built-in functions are
gradsf and gradsg for fixed and gaussian grids.

So it seems I need to either compute the divergence and gradient by
myself, or to convert the regular lat/lon grid onto a fixed cartesian
grid, and then use the accurate built-in functions uv2dvf and gradsf.
What would you advice to me? In case I need to compute the divergence
and gradient by myself, do you provide any source code of your built-
in functions that I could modify to apply the calculations on a
regular grid? Otherwise, what ncl code would you suggest to convert
the regular grid onto a fixed grid? Or is there another routine I
could use and that I didn't see?

The last thing is that I need then to integrate the divergence
vertically. My vertical levels are pressure levels, not equally
spaced: 1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100,
70, 50, 30, 20. What routine would you suggest to do that calculation?

Thank you very much in advance.

Best regards,
Marie

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Thu May 13 12:28:49 2010