Since movement in 2-D is a special case of movement in 3-D, the
eRTG3D algorithm also supports two-dimensional simulations. The underlying data structure of the algorithm remains in three-dimensional, with the third dimension (z) being constant, as for example zero. This approach guarantees a seamless transition between 2-D and 3-D simulations. Therefore, two P and Q probabilities are be extracted from 2D and 3D trajectories, then a combined simulation can take place.
To simulate in 2-D the third dimension of the trajectory is set to zero:
If the original trajectory is already two-dimensional, a third column
z has to be added:
trajectory.2D$z <- 0.
Now the workflow is the same as in 3-D, described in the standard workflow vignette:
Note: Since it is not feasible to use a DEM (
DEM = demRaster) in 2-D simualtions, the adding of a DEM in the somulations will result in dead ends. A BG layer (
BG = bgRaster) with a binary mask or continous probabilities for the simulation area can be passed (e.g. water bodies, nutrition sources, …).
And plotting the results: