A simpler and, maybe, safer, version of ieee2grb_sst_scoar.F

When first Dian and I changed the code, we decided to use subroutine setgrd within subroutine subgrid because we needed to get initial and final values of longitudes and latitudes. These values were not correctly obtained from the origianl ieee2grb_sst.F, because it assumed uniform dy, which is not true.
So we had to use setgrd to get flon and flat. The xlon1, xlon2, xlat1 and xlat2 (all in radian) are necessary to get the whole grid mesh in subroutine gribit within ieee2grb_sst.F.

The problem was that setgrid doen't work correctly in IBM machine and the resultant xlon2 and xlat2 were not correct. And I couldn't figure out why. Well. but since we need only the initial and final longitudes and latitudes of the RSM grid (which is already known!), I modified this code and let iee2grb_sst_scoar.F read these four values plus delx and dely before gribit.
See code.. ieee2grb_sst_scoar.F
The part I added in looks like this
! *******
! edit 11/2/2007
! hyodae
       open(101,file='fort.101',form='formatted')
!       read(101,*) llons,llone,llats,llate
       read(101,*) xlon1a, xlon2a, xlat1a, xlat2a, delx, dely
        ortru=30.
        idrt=1
! change degree to radian
       xlon1=xlon1a*(3.141593/180.)
       xlon2=xlon2a*(3.141593/180.)
       xlat1=xlat1a*(3.141593/180.)
       xlat2=xlat2a*(3.141593/180.)
       xlon11=xlon1
       xlat11=xlat1
!*********************
 
rcdt.dat was deleted and won't be used anymore.
rgdt.dat is modified and will be something like
114.240
245.040
18.489
65.001
105000
100000