Document Type : Research Paper

Author

Associate Professor of Urban planning, Malek-Ashtar University of Technolog

Abstract

Statistical analysis is an important tool for environmental planning, control and assessment. Most geographic information systems determine basic descriptive statistical indexes such as mode, mean, average, minimum/maximum and quantitative data of standard deviation, and describe characteristics that occur in a certain area, including determination of the average elevation in the area within a radius of five kilometers around a point or provision of diagrams of frequency which show the distribution of values from least to most within an area. Simple statistical comparison of two or more data layers such as cross tabulation can usually be carried out within the framework of database management system created in GIS, and complex methods such as multi-regression and analysis of variance can usually be implemented by transferring data files from GIS to a general or special statistical analysis software package which has a range of progressive analysis methods, and then transfer the results back to GIS.
GIS modeling is related to application of more complex mathematical methods for prediction of numerical values of environmental phenomena such as solar radiation or soil degradation at a particular point in the studied area. Models are based on simple linear equations which can be considered as a series of mathematical operations within the framework of GIS. Most of progressive models, including those based on solution of differential equations with several repetitions to reach the desirable solution, need a connection between GIS data and a suitable modeling program. The results are usually received in the form of new data layers that can be used in future analyses. These types of file transfer are necessary in any kind of analysis, because it is not reasonable to expect a geographic information system to have all modeling methods.