DINAMICA – A Landscape Dynamics Simulation

DINAMICA – A Landscape Dynamics Simulation Software
B. S. SOARES-FILHO1, A. DE A. ARAÚJO2, G. C. CERQUEIRA1, W. L. ARAÚJO1
{ CSR/IGC/UFMG, 2NPDI/DCC/UFMG}. Av. Antônio Carlos 6627, Belo Horizonte, MG.
{arnaldo, cerca}@dcc.ufmg.br, [email protected]
1
Abstract. This paper reports the development of a new spatial simulation model of landscape dynamics –
DINAMICA, which presents: 1) multi-scale vicinity-based transitional functions, 2) incorporation of spatial
feedback approach to a stochastic multi-step simulation engineering, and 3) the application of logistic
regression or weights of evidence to calculate the spatial dynamic transition probabilities. Application of
DINAMICA includes the prediction of a region's spatial pattern evolution according to pre-defined transition
rates.
1
Introduction
Simulation models can be envisaged as a heuristic device
useful to test hypothesis about landscape evolution under
several scenarios. As a result, several researchers have
dedicated themselves to the development of landscape
dynamics simulation models, thus contributing to a
diversity of approaches, which can be found in works such
as Turner [1], Dale et. al. [2], and Gilruth et al. [3]. This
paper presents the development of a landscape dynamics
software, DINAMICA, which is designed to simulate the
genesis and development of land change spatial patterns.
2
Software structure
DINAMICA software involves a multiple time step
stochastic simulation with dynamic spatial transition
probabilities
calculated
within
a
cartographic
neighborhood. Its engineering employs special functions
designed to reproduce cartographically the dimensions and
forms of landscape patches. For the model
parameterization, either logistic regression or weights of
evidence is applied to indicate the areas most favorable for
each type of transition.
The software uses as its main input a landscape map
(land-use and cover map) and employs selected spatial
variables which are structured in two cartographic subsets
according to their dynamic or static nature. It generates, as
output, simulated landscape maps (one for each time step),
the spatial transition probability maps, which depict the
probability of a cell at a position (x,y) to change from state
i to state j, and the dynamic spatial variable maps.
3
Application
DINAMICA software was applied to simulate the
landscape changes of an Amazonian frontier colonization
region, located in Northern Mato Grosso, Brazil. The time
span chosen for running the model encompassed eight
years and was divided into two periods: 1986 to 1991 and
1991 to 1994. The results from the validation methods
showed that the simulations were able to reproduce the
contagion indices and the fractal dimension; the multiple
resolution fitting procedure [4] showed an adjustment, at
highest resolution, from 63.6 to 82.4% for the reference
landscapes.
3
Conclusion
The software can be used to investigate several
environmental dynamic phenomena, considering the fact
that its transitional functions can be adapted to replicate
different landscape structures and to work with any spatial
resolution or at diverse cartographic scales. DINAMICA
can model any type and any number of transitions as well
as embrace any span of time, divided into any number of
time steps and phases with pre-defined transition rates.
The DINAMICA software were able to handle substantive
large data arrays with a good performance.
References
[1] Turner, M.G., 1987. Spatial simulation of landscape
changes in Georgia: a comparison of 3 transition models.
Landscape Ecology, 1: 27-39.
[2] Dale, V.H., O’Neill, R.V., Southworth, F., Pedlowski,
M., 1994. Modeling effects of land management in the
Brazilian
Amazonian
settlement
of
Rondônia.
Conservation Biology, 8: 196-206.
[3] Gilruth, P., Marsh, S.E., Itami, R., 1995. A dynamic
spatial model of shifting cultivation in the highlands of
Guinea, West Africa. Ecological Modelling, 79: 179-197.
[4] Turner, M.G., Costanza, R., Sklar, F., 1989. Methods
to evaluate the performance of spatial simulation models.
Ecological Modelling, 48: 1-18.
Proceedings of the XIV Brazilian Symposium on Computer Graphics and Image Processing (SIBGRAPI’01)
1530-1834/02 $17.00 © 2001 IEEE