The enhancement of a commercial geographical information system (ARC/INFO) with fuzzy processing capabilities for the evaluation of land resources

The paper presents the design considerations of FUZZYLAND; a geographical information system with fuzzy processing capabilities. The GIS package ARC/INFO was augmented with new commands to allow the easy application of fuzzy sets theory methods to geographical data. The system was programmed in AML (the ARC/INFO Macro Language) and operates as a set of ARC/INFO commands. This type of development makes the system very easy to use and allows complete control of every stage of the process. Examples of the operation of the FUZZYLAND system are presented, through a case study which demonstrates its application for the classification of the soils of an area of Western Greece in thematic classes according to their (a) degradation from agricultural uses and (b) to their suitability for growing alfalfa. The proposed system maintains all the advantages of the GIS ARC/INFO but enhances it in three ways: (1) it can easily manipulate the natural variation of soil properties; (2) users are allowed to express their subjective view on the stored data in the form of membership grades; and (3) it presents a more precise and realistic classification and assessment of land resources. (C) 1998 Elsevier Science B.V. All rights reserved.