Hi,
In other land cover modelling software, we can do simulations by specifying the surface needed for the simulated year and for each land type. I would like to do the same in LCM but we can't enter the demand surface.
If I got it right, I need to use the "External Model" option instead of the built-in Markov chain model to do that.
Do you have any ideas about which external model I should use in order to generate that kind of probabilty matrix based on surface demand ?
Hi,
If I understand correctly, you are referring to the demand surface, not the demand quantity. LCM uses a demand surface for the simulation. This is created when developing the model, one for each transition used in the model. We call these transition potential maps and they are created when yo are completed running one of the three modeling tools, MLP, SimWeight, or Logistic Regression.
James Toledano
Clark Labs
Thank you for your reply but I'm going to try to make it clearer.
For exemple, in the modelling software Dyna-CLUE, you can simulate different scenario. First, you write a storyline for the land cover developpement of your region and then you translate it into demand surface, area restriction, land use type elasticity...
The demand surface is the total surface for each land use type and for the simulated year. (For exemple i can say that for my region study i want 2000 ha of forest, 200 ha of water, 1000 ha of build-up... for the simulated land cover of the year 2050). These surface can be derived from trend extrapolation, from rough assumption (for exemple, an increase of 10% of the forest du to a high environmental protection policy) or ....
If I understand it correctly, in LCM, I can only obtain a simulation that follow the past trend when using the built-in Markov chain model. But i would like to be able to define my own storyline or scenario by defining the demand surface exactly like in Dyna-CLUE. I want to have control over the total surface for each land use type and for the simulated year.
Is it possible to achieve using the External Model option under the Change prediction of LCM ? If so, what kind of software should I use to create the right probability matrix ?
OK, what you are asking is a bit tricky, but doable in LCM. I think you are on the right track in that you will need to alter or create your own probability matrix. This is simply a text file, the format if which is on the Help.
The tricky part is defining the probabilities which are ultimately derived from the transition areas. This gets even more difficult the more transitions you have and if some of the land covers are transitioning both ways. There is a good discussion on how the area and probability matrices are derived in the Markov module Help Notes section. You use this same logic to create your own matrix. But you need to keep in mind that LCM wants a probability matrix for the external model, so you will need to convert those new areas into probabilities. Again, there is a nice discussion on how to do this in the Markov module Help section.
Best,
James Toledano
Clark Labs