LOCATION MODELS FOR MAIN INDUSTRIES AT MACRO-GEOGRAPHICAL LEVEL USING OPEN GEOSPATIAL DATA AND SOFTWARE: CASE OF ROMANIAN COUNTIES

Abstract

This paper reveals the mathematical models to explain the location of industries within macro-geographical areas (Romania’s counties), by means of GIS (Geographical Information System) metrics. While the explained variable (output) was a composite indicator equally weighting employment and turnover quotients computed from statistical data, the explaining variables (inputs) were a set of GIS metrics, computed on geospatial open data, using open GIS software. The preferred method for modelling was the multiple linear regression; different nonlinear functions were tested to provide the best fit. The GIS metrics are an alternative to statistical data, having the advantage of being procured and updated easier by the automatic import of the GIS database. For all industries, the study delivered relevant models. This study is part of a spatial decision support system (SDSS) for the location of enterprises, including both a macro-geographical layer and a micro-geographical layer of factors. This paper’s results can be used independently (the micro-geographical layer was addressed by a previous study). The macro-geographical and the micro-geographical layer could share the 100% weight according to the desired level of location. For example, a governmental organization could grant all weight to the macro layer, while an enterprise could weight more the micro layer.