C $Header: /home/ubuntu/mnt/e9_copy/MITgcm/verification/aim.5l_cs/code/SHAP_FILT_OPTIONS.h,v 1.4 2011/12/24 01:17:51 jmc Exp $ C $Name: $ C CPP options file for pkg SHAP_FILT #ifndef SHAP_FILT_OPTIONS_H #define SHAP_FILT_OPTIONS_H #include "PACKAGES_CONFIG.h" #include "CPP_OPTIONS.h" #ifdef ALLOW_SHAP_FILT C Package-specific Options & Macros go here C Use [1-d_yy^n)(1-d_xx^n] instead of [1-d_xx^n-d_yy^n] C This changes the spectral response function dramatically. C You need to do some analysis before changing this option. ;^) #define SEQUENTIAL_2D_SHAP C This invokes the older code which produces the same numbers C and might even be faster because it does not do multiple C exchanges within the filter. However, this requires the C overlap to be sufficiently wide and also does not work C for arbitrarily arranged tiles (ie. as in cubed-sphere). C *DO NOT USE THIS OPTION UNLESS YOU REALLY WANT TO* :-( #undef USE_OLD_SHAPIRO_FILTERS #ifdef USE_OLD_SHAPIRO_FILTERS C Boundary conditions for U,V Shapiro code C Horizontal shear is calculated as if the boundaries are no-slip C Note: option NO_SLIP_SHAP only used in OLD_SHAPIRO_FILTERS ; C it is replaced by parameter "Shap_noSlip=1." in new S/R. #undef NO_SLIP_SHAP #endif C Masking of vorticity in dissipation terms (and Shap-Filt) C can be different from the masking applied for momentum advection; C This option allows to use the local S/R: SHAP_FILT_RELVORT3 to C compute vorticity, instead of pkg/mom_common S/R: MOM_CALC_RELVORT3 #define USE_SHAP_CALC_VORTICITY #endif /* ALLOW_SHAP_FILT */ #endif /* SHAP_FILT_OPTIONS_H */