next up previous
Next: 6. Conclusions Up: 5. Standard Case Previous: 5.2 Temporal Variations of

5.3 Performance of the Compact Schemes

Feasibility of using the sixth order scheme is two-folds: 1) drastic error reduction, and 2) no drastic CPU time increase. We use four quantities to compare the errors: the maximum peak error velocity during the 10 day integration (Vtex2html_wrap_inline636), the averaged peak error velocity during 5-10 day integration (Vtex2html_wrap_inline638), the maximum peak error pressure gradient during the 10 day integration (PGtex2html_wrap_inline636), and the averaged peak error pressure gradient during 5-10 day integration (PGtex2html_wrap_inline638). Vtex2html_wrap_inline636 (or PGtex2html_wrap_inline636) indicates the maximum error during the geostrophic adjustment stage, and Vtex2html_wrap_inline638 (or PGtex2html_wrap_inline638) shows the mean error after the adjustment. Usually, Vtex2html_wrap_inline652 Vtex2html_wrap_inline638, and PGtex2html_wrap_inline652PGtex2html_wrap_inline658 The performance of each scheme is listed in Table 1.



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