Fix subvol for cell volume seeding

[oj-tooth]

When seeding trajectories using the nqua=3 case, the cell volume/mass is used for calculating the number of trajectories to be seeded:

! Number of trajectories for box (iist,ijst,ikst)
SELECT CASE (nqua)
...
CASE (3) ! particle reflects air/water mass/volume at seeding
                      vol = dzt(ib,jb,kb,1)
                      num = INT(vol/partQuant)
END SELECT

where subvol is determined using vol as follows:

subvol = vol / DBLE (ijt*ikt)

However, the current implementation assigns vol the vertical grid cell thickness at T-points rather than the grid cell volume.

Proposal
To assign vol as the grid cell volume to be distributed amongst the seeded trajectories:

! Number of trajectories for box (iist,ijst,ikst)
SELECT CASE (nqua)
...
CASE (3) ! particle reflects air/water mass/volume at seeding
                      vol = dxdy(ib,jb) * dzt(ib,jb,kb,nsp) * zstot(ib,jb,0)
                      num = INT(vol/partQuant)
END SELECT