| Resumen |
This paper describes landscape design metrics from the implementation of an Evolutionary Multi-objective
Optimization (EMO) algorithm for peri-urban greenlands system design. The results of particular landscape met-
rics using a real world data set of a small study area bring into relief issues concerning the interplay between the
mathematization of landscape ecology principles of design and the resulting set of estimates of the Pareto optimal
solutions. The paper discusses the research findings and provides links to other aspects of current EMO-based
research methodology and its potential use as part of a landscape design decision support system. Specifically, this
paper has two purposes. First, the design approach used may provide assistance to other researchers working to
develop metrics and associated objective functions that incorporate landscape feature shape and configuration into
decision support for environmental land-use planning. Second, the introduction of the new concept of anti-cut-set
vertices, derived from the graph theoretic concept of cut-set vertices, allows some key connections to be identified
between the larger sub-graphs in the dual graph of the study area’s landscape features. This gives a very direct
analytic tool for exploring potential conectivity structure in a given landscape |
