grad: Calculate the grad of a matrix by first differences

In oce: Analysis of Oceanographic Data

Description

In the interior of the matrix, centred second-order differences are used to infer the components of the grad. Along the edges, first-order differences are used.

Usage

grad(h, x, y)

Arguments

h	a matrix
x	x values
y	y values

Value

A list containing gx and gy, matrices of the same dimension as h.

Author(s)

Dan Kelley, based on advice of Clark Richards, and mimicking a matlab function.

See Also

Other functions relating to vector calculus: curl

Examples

## Geostrophic flow around an eddy
library(oce)
dx <- 5e3
dy <- 10e3
x <- seq(-200e3, 200e3, dx)
y <- seq(-200e3, 200e3, dy)
R <- 100e3
h <- outer(x, y, function(x, y) 500*exp(-(x^2+y^2)/R^2))
grad <- grad(h, x, y)
par(mfrow=c(2, 2), mar=c(3, 3, 1, 1), mgp=c(2, 0.7, 0))
contour(x,y,h,asp=1, main=expression(h))
f <- 1e-4
gprime <- 9.8 * 1 / 1024
u <- -(gprime / f) * grad$gy
v <-  (gprime / f) * grad$gx
contour(x, y, u, asp=1, main=expression(u))
contour(x, y, v, asp=1, main=expression(v))
contour(x, y, sqrt(u^2+v^2), asp=1, main=expression(speed))

oce documentation built on Aug. 29, 2017, 1:05 a.m.