The basis of the land use model is land use demand, sometimes called “land claims”. This is usually exogenous, in other words, it must be determined outside of the model. For scenario development, the demand for future dates needs to be considered carefully. For the calibration of the model, we already know the demand for the timestep T2, it’s just the land use in the existing map for T2 that we are going to use to compare against our simulation of the same date (in the case of this example, 1999). So, in R:
[1,] 0 255597
[2,] 1 6752
[3,] NA 599750
So we know that we have to locate 6752 cells between 1956 and 1999. But how many cells do we allocate per year?
Well, it’s quite likely that the new land use was not allocated evenly over the 43 time period, but we don’t have any more information, so we just have to do a simple linear interpolation between the two historic dates, thus (6752/43 = 157.023255814).
so we need to allocate 157 cells at each timestep!