- Fit a variogram model to the experimental variogram calculated from bottom trawl density data using the model.auto function:

- Code: Select all
`model.vario.Y.15 <- model.auto(vario.Y.15, struc=melem.name(c(1,2)), draw=T)`

- Initiate a grid and neighbourhood

- Code: Select all
`grid.simu <- db.grid.init(poly.data,nodes=c(gnx,gny))`

neigh.simu <- neigh.init(ndim=2,type=2,nmini=5,nmaxi=100,dmax=60)

- Simulate across the grid

- Code: Select all
`grid.simu <- simtub(dbin=db.data.trans, dbout=grid.simu, model=model.vario.Y.15,`

neigh=neigh.simu, uc="", mean=0, seed=1337, nbsimu=ns,

nbtuba=1000, radix="Simu", modify.target=TRUE)

In most cases the entire grid (which is based on an outline polygon of the area of interest) is filled with simulated values. In some cases though, the grid does not fill and the extent of the simulated surface appears to be limited to some of the data points.

Is this related to the amount of data points, the variograms properties, or the simulation itself?