next up previous contents
Next: 31.1.2 Leap frog in Up: 31.1 Considerations of accuracy Previous: 31.1 Considerations of accuracy

   
31.1.1 Lattice and continuum operators

A good reference for the methods used in this subsection is chapter 2 from the book by Mitchell and Griffiths (1980).

Consider the continuum field $\psi$ evaluated at an arbitrary lattice point

$\displaystyle \psi^{n}_{i} = \psi(x_{i},t_{n}).$     (31.2)

Using a Taylor series expansion, the relation between $\psi^{n}_{i}$and $\psi^{n}_{i + p}$, where $x_{p} = p \, \Delta x$, is given by
$\displaystyle \psi^{n}_{i + p}$ = $\displaystyle \left( 1 + p \, \Delta x \, \partial_{x} + \frac{1}{2!} \,
(p \, ...
...{1}{3!} \, (p \, \Delta x)^{3} \partial^{(3)}_{x}
+ \ldots \right) \psi^{n}_{i}$  
  = $\displaystyle \exp( p \, \Delta x \, \partial_{x}) \, \psi^{n}_{i}.$ (31.3)

The linear operator $\exp( p \, \Delta x \, \partial_{x})$ can be thought of as a spatial translation operator; a terminology familiar to those having studied quantum mechanics. Similarly, the temporal translation operator $\exp(q \, \Delta t \, \partial_{t})$ connects $\psi^{n}_{i}$ to another point in time $t_{q} = q \, \Delta t$through the relation
$\displaystyle \psi^{n + q}_{i} = \exp(q \, \Delta t \, \partial_{t}) \, \psi^{n}_{i}.$     (31.4)

In the following, it will prove useful to derive relations between the continuum differential operators $\partial_{x}$ and $\partial_{t}$ and various finite difference or lattice operators. To start, consider the familiar centered difference spatial operator as defined by
$\displaystyle \delta^{C}_{x} \psi_{i}^{n}
= \frac{\psi_{i+1}^{n} -\psi_{i-1}^{n}}{2 \Delta x}.$     (31.5)

Using the spatial translation operators, this relation takes the form
$\displaystyle \delta^{C}_{x} \psi_{i}^{n}
= \left( \frac{\sinh(\Delta x \, \partial_{x}) }{\Delta x} \right) \psi_{i}^{n}.$     (31.6)

Since $\psi_{i}^{n}$ is arbitrary, this equation yields the relation between the centered difference lattice operator and the continuous partial derivative
$\displaystyle \Delta x \, \delta_{x}^{C} = \sinh(\Delta x \, \partial_{x}).$     (31.7)

Inverting this relation yields
$\displaystyle \Delta x \, \partial_{x} = \sinh^{-1}(\Delta x \, \delta^{C}_{x} ).$     (31.8)

Similar relations hold for the temporal centered difference, or leap frog, operator
$\displaystyle \Delta t \, \delta_{t}^{C}$ = $\displaystyle \sinh(\Delta t \, \partial_{t})$ (31.9)
$\displaystyle {\Delta t} \, \partial_{t}$ = $\displaystyle \sinh^{-1}(\Delta t \, \delta^{C}_{t} ).$ (31.10)

The forward difference lattice operator is also quite common

$\displaystyle \delta^{F}_{x} \psi_{i}^{n}$ = $\displaystyle \frac{\psi_{i+1}^{n} -\psi_{i}^{n}}{\Delta x}$  
  = $\displaystyle [ \exp(\Delta x \, \partial_{x}) - 1 ] \, \psi_{i}^{n}.$ (31.11)

This relation leads to the operator equalities
$\displaystyle \Delta x \, \delta^{F}_{x}$ = $\displaystyle \exp(\Delta x \, \partial_{x}) - 1$ (31.12)
$\displaystyle \Delta x \, \partial_{x}$ = $\displaystyle \ln(1 + \Delta x \, \delta_{x}^{F}),$ (31.13)

with analogous results for the temporal forward difference operator. The relation between other finite difference operators and the continuum operator can be derived similarly.


next up previous contents
Next: 31.1.2 Leap frog in Up: 31.1 Considerations of accuracy Previous: 31.1 Considerations of accuracy
RC Pacanowski and SM Griffies, GFDL, Jan 2000