next up previous
Next: 2. Compact Difference Schemes Up: Improvement of Estuarine and Previous: Abstract

1. Introduction

Most estuarine and coastal models use second-order difference schemes (such as second-order staggered C-grid scheme) to approximate first-order derivative (Blumberg and Mellor, 1987; Hadivogel et al., 1991)
equation57
where p, tex2html_wrap_inline536 represent pressure and grid spacing. Such a difference scheme was proposed by numerical modelers in early 50's as the first generation computers came into place. Since then the computer updates rapidly with several orders of magnitude increase in computational power. However, the difference schemes used by most modelers now are still staying at the 50's level (second-order schemes).

If we go a step further by using the fourth-order C-grid scheme (Figure 1a) proposed by McCalpin [1994],
equation69
the truncation error is on the order of O(tex2html_wrap_inline538). Comparing (2) with (1), the error ratio between the fourth-order and second-order schemes is estimated by
equation85
which is proportional to tex2html_wrap_inline540

Since the truncation error decreases with the increase of the order of the difference scheme, it might be benefited to use an even higher order difference scheme. Chu and Fan [1997] proposed a sixth-order difference scheme
displaymath532

equation115
to compute the horizontal pressure gradient. Comparing (4) with (1), the error ratio between the sixth-order and second-order schemes is estimated by
equation122
which is proportional to tex2html_wrap_inline542

There are two weaknesses of using ordinary high-order difference schemes. The first one is more grid points needed for the computation: the ordinary 4-th order scheme (2) needs 4 points and the ordinary 6-th order scheme (4) needs 6 points. The second one is using lower-order schemes at two boundaries. Taking the fourth-order scheme as an example, gradient at the 'left' boundary has second-order accuracy
equation142


Peter Chu
Thu Aug 24 17:01:43 PDT 2000