sufctanismod.su.main

su 的源代码库

 SUFCTANISMOD - Flux-Corrected Transport correction applied to the 2D
	  elastic wave equation for finite difference modeling in 	
	  anisotropic media						

 sufctanismod > outfile [optional parameters]			
		outfile is the wavefield snapshot x-component		
		x-component of wavefield snapshot is in snapshotx.data	
		y-component of wavefield snapshot is in snapshoty.data	
		z-component of wavefield snapshot is in snapshotz.data	

 Optional Output Files:						
 reflxfile=	reflection seismogram file name for x-component		
		no output produced if no name specified	 		
 reflyfile=	reflection seismogram file name for y-component		
		no output produced if no name specified	 		
 reflzfile=	reflection seismogram file name for z-component		
		no output produced if no name specified	 		
 vspxfile=	VSP seismogram file name for x-component		
		no output produced if no name specified	 		
 vspyfile=	VSP seismogram file name for y-component		
		no output produced if no name specified	 		
 vspzfile=	VSP seismogram file name for z-component		
		no output produced if no name specified	 		

 suhead=1      To get SU-header output seismograms (else suhead=0)	

 New parameter:							
 receiverdepth=0  depth of horizontal receivers (in gridpoints)      

     
 Optional Parameters:						
 dofct=1 	1 do the FCT correction					
		0 do not do the FCT correction 				
 FCT Related parameters:						
 eta0=0.03	diffusion coefficient					
		typical values ranging from 0.008 to 0.06		
		about 0.03 for the second-order method 			
		about 0.012 for the fourth-order method 		
 eta=0.04	anti-diffusion coefficient 				
		typical values ranging from 0.008 to 0.06		
		about 0.04 for the second-order method  		
		about 0.015 for the fourth-order method 		
 fctxbeg=0 	x coordinate to begin applying the FCT correction	
 fctzbeg=0 	z coordinate to begin applying the FCT correction	
 fctxend=nx 	x coordinate to stop applying the FCT correction	
 fctzend=nz 	z coordinate to stop applying the FCT correction	

 deta0dx=0.0	gradient of eta0 in x-direction  d(eta0)/dx		
 deta0dz=0.0	gradient of eta0 in z-direction  d(eta0)/dz		
 detadx=0.0	gradient of eta in x-direction 	 d(eta)/dx		
 detadz=0.0	gradient of eta in z-direction 	 d(eta)/dz		

 General Parameters:							
 order=2	2 second-order finite-difference 			
		4 fourth-order finite-difference 			

 nt=200        number of time steps 			 		
 dt=0.004	time step  						

 nx=100 	number of grid points in x-direction 			
 nz=100 	number of grid points in z-direction 			

 dx=0.02	spatial step in x-direction 				
 dz=0.02	spatial step in z-direction 				

 sx=nx/2	source x-coordinate (in gridpoints)			
 sy=nz/2	source z-coordinate (in gridpoints)			

 fpeak=20	peak frequency of the wavelet 				

 wavelet=1	1 AKB wavelet						
 		2 Ricker wavelet 					
		3 impulse 						
		4 unity 						

 isurf=2	1 absorbing surface condition 				
		2 free surface condition 				
		3 zero surface condition 				

 source=1	1 point source 						
 		2 sources are located on a given refelector 	        ", 
			(two horizontal and one dipping reflectors) 	
 		3 sources are located on a given dipping refelector     ", 

 sfile= 	the name of input source file, if no name specified then
		use default source location. (source=1 or 2) 		

 Density and Elastic Parameters:					
 dfile= 	the name of input density file,                         
               if no name specified then                             
		assume a linear density profile with ...		
 rho00=2.0	density at (0, 0) 					
 drhodx=0.0	density gradient in x-direction  d(rho)/dx		
 drhodz=0.0	density gradient in z-direction  d(rho)/dz		

 afile= 	name of input elastic param.  (c11) aa file, if no name 
		specified then, assume a linear profile with ...	
 aa00=2.0	elastic parameter at (0, 0) 				
 daadx=0.0	parameter gradient in x-direction  d(aa)/dx		
 daadz=0.0	parameter gradient in z-direction  d(aa)/dz		

 cfile= 	name of input elastic param. (c33)  cc file, if no name 
		specified then, assume a linear profile with ...	
 cc00=2.0	elastic parameter at (0, 0) 				
 dccdx=0.0	parameter gradient in x-direction  d(cc)/dx		
 dccdz=0.0	parameter gradient in z-direction  d(cc)/dz		

 ffile= 	name of input elastic param.  (c13) ff file, if no name 
		specified then, assume a linear profile with ...	
 ff00=2.0	elastic parameter at (0, 0) 				
 dffdx=0.0	parameter gradient in x-direction  d(ff)/dx		
 dffdz=0.0	parameter gradient in z-direction  d(ff)/dz		

 lfile= 	name of input elastic param.  (c44) ll file, if no name 
		specified then, assume a linear profile with ...	
 ll00=2.0	elastic parameter at (0, 0) 				
 dlldx=0.0	parameter gradient in x-direction  d(ll)/dx		
 dlldz=0.0	parameter gradient in z-direction  d(ll)/dz		

 nfile= 	name of input elastic param. (c66)  nn file, if no name 
		specified then, assume a linear profile with ...	
 nn00=2.0	elastic parameter at (0, 0) 				
 dnndx=0.0	parameter gradient in x-direction  d(nn)/dx		
 dnndz=0.0	parameter gradient in z-direction  d(nn)/dz		

 Optimizations:							
 The moving boundary option permits the user to restrict the computations
 of the wavefield to be confined to a specific range of spatial coordinates.
 The boundary of this restricted area moves with the wavefield		
 movebc=0	0 do not use moving boundary optimization		
		1 use moving boundaries					
  If movebc=1 then specify:						
 mbx1=0 	initial left side of moving boundary			
 mbz1=0 	initial top of moving boundary 				
 mbx2=nx 	initial right side of moving boundary 			
 mbz2=nz 	initial bottom of moving boundary 			



 Author: Tong Fei,	Center for Wave Phenomena, 
		Colorado School of Mines, Dec 1993
 Some additional features by: Stig-Kyrre Foss, CWP
		Colorado School of Mines, Oct 2001
 New features (Oct 2001): 
 - setting receiver depth
 - outputfiles with SU-headers
 - additional commentary


 
Notes:
	This program performs seismic modeling for elastic anisotropic 
	media with vertical axis of symmetry.  
	The finite-difference method with the FCT correction is used.

	Stability condition:	vmax*dt /(sqrt(2)*min(dx,dz)) < 1
	
	Two major stages are used in the algorithm:
	(1) conventional finite-difference wave extrapolation
	(2) followed by an FCT correction 

References:
	The detailed algorithm description is given in the article
	"Elimination of dispersion in finite-difference modeling 
	and migration"	in CWP-137, project review, page 155-174.

	Original reference to the FCT method:
	Boris, J., and Book, D., 1973, Flux-corrected transport. I.
	SHASTA, a fluid transport algorithm that works: 
	Journal of Computational Physics, vol. 11, p. 38-69.

/