| 1 |
C $Header: /u/gcmpack/MITgcm/pkg/flt/flt_runga2.F,v 1.5 2004/09/13 16:13:42 edhill Exp $ |
| 2 |
C $Name: $ |
| 3 |
|
| 4 |
#include "FLT_CPPOPTIONS.h" |
| 5 |
|
| 6 |
subroutine flt_runga2 ( |
| 7 |
I myCurrentIter, |
| 8 |
I myCurrentTime, |
| 9 |
I myThid |
| 10 |
& ) |
| 11 |
|
| 12 |
c ================================================================== |
| 13 |
c SUBROUTINE flt_runga2 |
| 14 |
c ================================================================== |
| 15 |
c |
| 16 |
c o This routine steps floats forward with second order Runge-Kutta |
| 17 |
c |
| 18 |
c started: Arne Biastoch |
| 19 |
c |
| 20 |
c changed: 2004.06.10 Antti Westerlund (antti.westerlund@helsinki.fi) |
| 21 |
c and Sergio Jaramillo (sju@eos.ubc.ca) |
| 22 |
c |
| 23 |
c ================================================================== |
| 24 |
c SUBROUTINE flt_runga2 |
| 25 |
c ================================================================== |
| 26 |
|
| 27 |
c == global variables == |
| 28 |
|
| 29 |
#include "EEPARAMS.h" |
| 30 |
#include "SIZE.h" |
| 31 |
#include "DYNVARS.h" |
| 32 |
#include "PARAMS.h" |
| 33 |
#include "GRID.h" |
| 34 |
#include "FLT.h" |
| 35 |
#ifdef ALLOW_3D_FLT |
| 36 |
#include "GW.h" |
| 37 |
#endif |
| 38 |
|
| 39 |
c == routine arguments == |
| 40 |
|
| 41 |
INTEGER myCurrentIter, myThid |
| 42 |
_RL myCurrentTime |
| 43 |
INTEGER bi, bj |
| 44 |
_RL global2local_i |
| 45 |
_RL global2local_j |
| 46 |
_RL global2local_k |
| 47 |
c == local variables == |
| 48 |
|
| 49 |
integer ip, kp, iG, jG |
| 50 |
_RL phi, uu, vv, u1, v1 |
| 51 |
#ifdef ALLOW_3D_FLT |
| 52 |
_RL ww, w1, zt, zz, scalez |
| 53 |
#endif |
| 54 |
_RL xx, yy, xt, yt |
| 55 |
_RL scalex, scaley |
| 56 |
character*(max_len_mbuf) msgbuf |
| 57 |
_RL npart_dist |
| 58 |
#ifdef USE_FLT_ALT_NOISE |
| 59 |
Real*8 PORT_RAND_NORM |
| 60 |
#else |
| 61 |
Real*8 PORT_RAND |
| 62 |
#undef _USE_INTEGERS |
| 63 |
#ifdef _USE_INTEGERS |
| 64 |
integer seed |
| 65 |
seed = -1 |
| 66 |
#else |
| 67 |
Real*8 seed |
| 68 |
seed = -1.d0 |
| 69 |
#endif |
| 70 |
#endif |
| 71 |
|
| 72 |
c == end of interface == |
| 73 |
|
| 74 |
DO bj=myByLo(myThid),myByHi(myThid) |
| 75 |
DO bi=myBxLo(myThid),myBxHi(myThid) |
| 76 |
do ip=1,npart_tile(bi,bj) |
| 77 |
|
| 78 |
c If float has died move to level 0 |
| 79 |
c |
| 80 |
if( |
| 81 |
& (tend(ip,bi,bj).ne.-1. .and. myCurrentTime.gt. tend(ip,bi,bj))) |
| 82 |
& then |
| 83 |
kpart(ip,bi,bj) = 0. |
| 84 |
else |
| 85 |
c Start integration between tstart and tend (individual for each float) |
| 86 |
c |
| 87 |
if( |
| 88 |
& (tstart(ip,bi,bj).eq.-1. .or. myCurrentTime.ge.tstart(ip,bi,bj)) |
| 89 |
& .and. |
| 90 |
& ( tend(ip,bi,bj).eq.-1. .or. myCurrentTime.le. tend(ip,bi,bj)) |
| 91 |
& .and. |
| 92 |
& ( iup(ip,bi,bj).ne. -3.) |
| 93 |
& ) then |
| 94 |
|
| 95 |
c Convert to local indices |
| 96 |
c |
| 97 |
|
| 98 |
C Note: global2local_i and global2local_j use delX and delY. |
| 99 |
C This may be a problem, especially if you are using a curvilinear |
| 100 |
C grid. More information below. |
| 101 |
xx=global2local_i(xpart(ip,bi,bj),bi,bj,mythid) |
| 102 |
yy=global2local_j(ypart(ip,bi,bj),bi,bj,mythid) |
| 103 |
kp=INT(kpart(ip,bi,bj)) |
| 104 |
|
| 105 |
scalex=recip_dxF(INT(xx),INT(yy),bi,bj) |
| 106 |
scaley=recip_dyF(INT(xx),INT(yy),bi,bj) |
| 107 |
iG = myXGlobalLo + (bi-1)*sNx |
| 108 |
jG = myYGlobalLo + (bj-1)*sNy |
| 109 |
|
| 110 |
|
| 111 |
#ifdef ALLOW_3D_FLT |
| 112 |
if (iup(ip,bi,bj).eq.-1.) then |
| 113 |
c zz=global2local_k(kpart(ip,bi,bj),bi,bj,mythid) |
| 114 |
|
| 115 |
c recip_drF is in units 1/r (so if r is in m this is in 1/m) |
| 116 |
scalez=recip_drF(kp) |
| 117 |
c We should not do any special conversions for zz, since flt_trilinear |
| 118 |
c expects it to be just a normal kpart type variable. |
| 119 |
zz=kpart(ip,bi,bj) |
| 120 |
call flt_trilinear(xx,yy,zz,uu,uVel,2,bi,bj) |
| 121 |
call flt_trilinear(xx,yy,zz,vv,vVel,3,bi,bj) |
| 122 |
call flt_trilinear(zz,yy,zz,ww,wVel,4,bi,bj) |
| 123 |
zt=zz+0.5*deltaTmom*ww*scalez |
| 124 |
else |
| 125 |
#endif |
| 126 |
call flt_bilinear(xx,yy,uu,kp,uVel,2,bi,bj) |
| 127 |
call flt_bilinear(xx,yy,vv,kp,vVel,3,bi,bj) |
| 128 |
#ifdef ALLOW_3D_FLT |
| 129 |
endif |
| 130 |
#endif |
| 131 |
|
| 132 |
#ifdef USE_FLT_ALT_NOISE |
| 133 |
c When using this alternative scheme the noise probably should not be added twice. |
| 134 |
#else |
| 135 |
if (iup(ip,bi,bj).ne.-2.) then |
| 136 |
uu = uu + uu*(PORT_RAND(seed)-0.5)*flt_noise |
| 137 |
vv = vv + vv*(PORT_RAND(seed)-0.5)*flt_noise |
| 138 |
#ifdef ALLOW_3D_FLT |
| 139 |
#ifdef ALLOW_FLT_3D_NOISE |
| 140 |
if (iup(ip,bi,bj).eq.-1.) then |
| 141 |
ww = ww + ww*(PORT_RAND(seed)-0.5)*flt_noise |
| 142 |
endif |
| 143 |
#endif |
| 144 |
#endif |
| 145 |
endif |
| 146 |
#endif |
| 147 |
|
| 148 |
c xx and xt are in indices. Therefore it is necessary to multiply |
| 149 |
c with a grid scale factor. |
| 150 |
c |
| 151 |
xt=xx+0.5*deltaTmom*uu*scalex |
| 152 |
yt=yy+0.5*deltaTmom*vv*scaley |
| 153 |
|
| 154 |
c Second step |
| 155 |
c |
| 156 |
|
| 157 |
#ifdef ALLOW_3D_FLT |
| 158 |
if (iup(ip,bi,bj).eq.-1.) then |
| 159 |
call flt_trilinear(xt,yt,zt,u1,uVel,2,bi,bj) |
| 160 |
call flt_trilinear(xt,yt,zt,v1,vVel,3,bi,bj) |
| 161 |
call flt_trilinear(xt,yt,zt,w1,wVel,4,bi,bj) |
| 162 |
else |
| 163 |
#endif |
| 164 |
call flt_bilinear(xt,yt,u1,kp,uVel,2,bi,bj) |
| 165 |
call flt_bilinear(xt,yt,v1,kp,vVel,3,bi,bj) |
| 166 |
#ifdef ALLOW_3D_FLT |
| 167 |
endif |
| 168 |
#endif |
| 169 |
|
| 170 |
if (iup(ip,bi,bj).ne.-2.) then |
| 171 |
#ifdef USE_FLT_ALT_NOISE |
| 172 |
u1 = u1 + port_rand_norm()*flt_noise |
| 173 |
v1 = v1 + port_rand_norm()*flt_noise |
| 174 |
#ifdef ALLOW_3D_FLT |
| 175 |
#ifdef ALLOW_FLT_3D_NOISE |
| 176 |
if (iup(ip,bi,bj).eq.-1.) then |
| 177 |
w1 = w1 + port_rand_norm()*flt_noise |
| 178 |
endif |
| 179 |
#endif |
| 180 |
#endif |
| 181 |
|
| 182 |
#else |
| 183 |
u1 = u1 + u1*(PORT_RAND(seed)-0.5)*flt_noise |
| 184 |
v1 = v1 + v1*(PORT_RAND(seed)-0.5)*flt_noise |
| 185 |
#ifdef ALLOW_3D_FLT |
| 186 |
#ifdef ALLOW_FLT_3D_NOISE |
| 187 |
if (iup(ip,bi,bj).eq.-1.) then |
| 188 |
w1 = w1 + w1*(PORT_RAND(seed)-0.5)*flt_noise |
| 189 |
endif |
| 190 |
#endif |
| 191 |
#endif |
| 192 |
|
| 193 |
#endif |
| 194 |
endif |
| 195 |
|
| 196 |
c xpart is in coordinates. Therefore it is necessary to multiply |
| 197 |
c with a grid scale factor divided by the number grid points per |
| 198 |
c geographical coordinate. |
| 199 |
c |
| 200 |
C This will only work if delX & delY are available. |
| 201 |
C This may be a problem, especially if you are using a curvilinear |
| 202 |
C grid. In that case you have to replace them for the values of |
| 203 |
C your grid, which can be troublesome. |
| 204 |
xpart(ip,bi,bj) = xpart(ip,bi,bj) |
| 205 |
& + deltaTmom*u1*scalex*delX(iG) |
| 206 |
ypart(ip,bi,bj) = ypart(ip,bi,bj) |
| 207 |
& + deltaTmom*v1*scaley*delY(jG) |
| 208 |
#ifdef ALLOW_3D_FLT |
| 209 |
if (iup(ip,bi,bj).eq.-1.) then |
| 210 |
kpart(ip,bi,bj) = kpart(ip,bi,bj) |
| 211 |
& + deltaTmom*w1*scalez |
| 212 |
endif |
| 213 |
#endif |
| 214 |
|
| 215 |
#ifdef ALLOW_3D_FLT |
| 216 |
c If float is 3D, make sure that it remains in water |
| 217 |
if (iup(ip,bi,bj).eq.-1.) then |
| 218 |
c reflect on surface |
| 219 |
if(kpart(ip,bi,bj).lt.1.0) |
| 220 |
& kpart(ip,bi,bj)=1.0 |
| 221 |
& +abs(1.0-kpart(ip,bi,bj)) |
| 222 |
c stop at bottom |
| 223 |
if(kpart(ip,bi,bj).gt.Nr) |
| 224 |
& kpart(ip,bi,bj)=Nr |
| 225 |
endif |
| 226 |
#endif |
| 227 |
endif |
| 228 |
endif |
| 229 |
|
| 230 |
enddo |
| 231 |
ENDDO |
| 232 |
ENDDO |
| 233 |
c |
| 234 |
return |
| 235 |
end |
Parent Directory
| 
Revision Log
| 
Revision Graph