Generating a Zoning Map
In the example data, zoning has been included simply as a No Data mask to all of the other maps. This means that no change will take place in the excluded area (In this case, the Doñana natural protected area). Once the simulation has been run, we just need to carry out a simple GIS map overlay operation to fill all the no data areas which overlap data in an unmasked map for the date of the initial map.
However, zoning can also be added as a separate map, and multiplied into the Transition Potential Computation; of course, setting the zoned areas to 0 seems problematic, because the TP computation is a product, and multiplying by zero will mean that the zoned areas will be filled with plain zeroes, not the land use in the first map. If we set zoned areas to 1, they will just take the TP score from the product of the other parameters, N (Neighbourhood), V (Stochastic factor), S (Suitability), and A (Accessibility), which isn’t right either, as these areas will still have some potential for change.
So what do we do?
We’re working on this, but it seem likely that we need a simple ‘if’ condition; e.g. if zoning=0, then t1=t1, in other words, the map is not updated where zoning is set to zero, reflecting the effect of a strict protected area restriction.
When we have a simple example worked out, we’ll post it here. But in the meantime the model works fine using the simple workaround described above.