Geography Reference
In-Depth Information
On the whole, the land use change simulation models can be broadly divided into
three major categories: empirical statistical model, agent-based model (ABM), and
raster neighborhood relationship-based model (Liu et al.
2005a
,
b
).
There are abundant empirical statistical models applied in land use change
simulation. This kind of models can be broadly divided into two categories:
econometric model that describes the process of land use change by establishing
equations between land use and its influencing factors, and mechanism model
identifying the relationship of land use change and its driving factors at grid levels.
A typical example of the latter is the Conversion of Land Use and its Effect at Small
regional extent (CLUE-S) model whose application in land use change simulation is
currently in the ascendant (Veldkamp and Fresco
1996
). The CLUE-S model is
constructed to simulate land use change and its effects on environment at meso-
micro scale. It has the capability of synchronously simulating the changes of mul-
tiple types and introduces the dynamic driving factors (such as population and
economic growth) to improve the simulation accuracy.
Since the 1990s, along with the rapid development of complexity science, ABM
began to be applied in land use change research. The Agent-based Models of Land
Use and Cover Change (ABM/LUCC) was specially discussed by LUCC Report
No.6, in which the development prospect of ABM in land use change simulation is
highly valued (McConnell
2001
). The ABM can be divided into two categories.
One is simulation model of landscape scale mainly based on traditional spatial
modeling techniques and the other depicts human decision-making processes and
their interactions (Semboloni et al.
2004
; Zhang et al.
2013
). The latter mainly
identifies the linkage between agents and environment by describing the interac-
tion and affiliation of independent agents (Manson
2006
). It is found that the
agents would get more benefits under the scenario without climate changes in the
long term, even though the total income is lower than that of under the scenario
with climate changes. Studies showed that ABM is efficient in describing the
interaction between macro individual and micro individual.
As a representative of raster neighborhood relationship-based model, CA model
is widely used in land use change simulation, especially urban expansion. Syphard
et al. (
2005
) analyzed the distinction of LUCC caused by urban expansion in areas
with different slope with the CA model. One of the superiority of the CA model in
land use change simulation is that it supports visualization of the simulation
process. The structure of the CA model makes it difficult considering the impacts
of land use policies. By combining ABM and Cellular Automata (CA) model, the
simulation of land use change is characterized by multi-scale and becomes more
effective in multi-objective decision making.
The existing models including CLUE-S model, ABM, and CA model are not
compatible with RCMs. The CLUE-S model needs an input of the land use
structure and ignores the influences of climate zone change on land use change.
The ABM and CA model is good at urban expansion simulation but vegetation
change driven by natural environment condition change. In this study, we devel-
oped a LUCD model in compatible with RCMs to describe the interdependencies
and feedback mechanisms among social economics, ecosystem environment as
