spgradq.f File Reference

Compute gradient in spectral space. More...

Go to the source code of this file.

Functions/Subroutines
subroutine	spgradq (i, m, enn1, elonn1, eon, eontop, q, qdx, qdy, qdytop)
	Computes the horizontal vector gradient of a scalar field in spectral space.

Detailed Description

Compute gradient in spectral space.

Author

Iredell

Date

92-10-31

Definition in file spgradq.f.

Function/Subroutine Documentation

spgradq()

subroutine spgradq	(		i,
			m,
		real, dimension((m+1)*((i+1)*m+2)/2)	enn1,
		real, dimension((m+1)*((i+1)*m+2)/2)	elonn1,
		real, dimension((m+1)*((i+1)*m+2)/2)	eon,
		real, dimension(m+1)	eontop,
		real, dimension((m+1)*((i+1)*m+2))	q,
		real, dimension((m+1)*((i+1)*m+2))	qdx,
		real, dimension((m+1)*((i+1)*m+2))	qdy,
		real, dimension(2*(m+1))	qdytop
	)

Computes the horizontal vector gradient of a scalar field in spectral space.

Subprogram speps() should be called already.

If l is the zonal wavenumber, n is the total wavenumber, eps(l,n)=sqrt((n**2-l**2)/(4*n**2-1)) and a is earth radius, then the zonal gradient of q(l,n) is simply i*l/a*q(l,n) while the meridional gradient of q(l,n) is computed as eps(l,n+1)*(n+2)/a*q(l,n+1)-eps(l,n+1)*(n-1)/a*q(l,n-1).

Extra terms are computed over top of the spectral domain.

Advantage is taken of the fact that eps(l,l)=0 in order to vectorize over the entire spectral domain.

Parameters

I	spectral domain shape (0 for triangular, 1 for rhomboidal)
M	spectral truncation
ENN1
ELONN1
EON	EPSILON/N*A
EONTOP	EPSILON/N*A over top
Q	scalar field
QDX	zonal gradient (times coslat)
QDY	merid gradient (times coslat)
QDYTOP	merid gradient (times coslat) over top

Author

IREDELL

Date

92-10-31

Definition at line 33 of file spgradq.f.