Some questions about landscape modlling

ＭｏｄｅｌｌｉｎｇｅｃｏｓｙｓｔｅｍｓｈａｓｂｅｅｎｄｏｎｅｂｙｍａｎｙｐｅｏｐｌｅｓｉｎｃｅｔｈｅｉｎｉｔｉａｌｌａｕｎｃｈｇｉｖｅｎｂｙＯｄｕｍ (Ｏｄｕｍ ,1983) .Ｂｕｔｍｏｄｅｌｌｉｎｇｔｈｅｆｕｎｃｔｉｏｎｉｎｇｏｆｌａｎｄｓｃａｐｅｓｉｓｎｏｔｓｏｆｒｅｑｕｅｎｔａｎｄｌａｎｄｓｃａｐｅｅｃｏｌｏｇｙｉｓｓｏｙｏｕｎｇｔｈａｔｔｈｅｔｙｐｅｏｆｍｏｄｅｌｗｈｉｃｈｈａｓｔｏｂｅｃｈｏｓｅｎｉｓｎｏｔａｌｗａｙｓｅｖｉｄｅｎｔ.Ａｆｉｒｓｔｃｏｎｓｔｒａｉｎｔｉｓｔｈａｔｔｈｅｍｏｄｅｌｍ…

Some questions about landscape modelling

The paper discusses mainly about the modelling process and related problems with examples from Chinese and French cases. Five practical problems must be solved for modelling the functioning of any landscape: (1) The field data are necessarily taken with a sampling procedure that implies a spatial (and often temporal) scale. (2) Every landscape modelled has to be identified, delimited and characterised before application of the hierarchical theory. (3) The functioning of a landscape involves data of multiple types (climate, soil, vegetation, fauna, buildings, communications, economy, aesthetics, etc.) which must be integrated in a holistic approach. (4) Every landscape is spatially heterogeneous, and the structure of the model must be more or less isomorphic with its heterogeneity. (5) The evolution of the landscape must be modelled on a rather long period of time. For all these reasons, it is necessary to build ad hoc models. Object|oriented computing languages may be useful for this purpose.