Hi Minoo! In Landscape Pattern Analysis one of the most important topic is the landcover/landuse change! A landscape can be described by two fundamental characteristics: composition and configuration. Composition means the description and classification of the different cover types which form the landscape. Configuration means how the cover type are arranged into the landscape.

Usually, these type of analysis are done through landcover map comparison between two or more dates, performing a transition matrix, in which, for each land cover/land use type you can quantify the magnitude of change. In addition, landscape pattern indices (LPIs) are a good tool to describe the configuration of a landscape and monitor their change. This type of approach (LPIs) requires attention and good knowledge about the LPIs. In fact, LPIs are sensitive to the scale of the analysis (both grain and extent), miss classifications, minimum mapping unit, leves of aggregation or fragmentation, ecc... Moreover, a huge number of LPIs exists and many of them are redundant.

Hi Ludovico, thanks a lot for reseanble answer, I could gain the enough information of your answer, good luck

Right: you can use Markov Chains in order to compare landcover map at different time then you can use a Cellular Automata in order to evaluate some external variables (slope, distance from roads, etc).