Food industry site selection using geospatial technology approach

Abstract

Food security has been an ongoing concern of governments and international organizations. One of the main issues in food security in Developing and Sanctioned Countries (DSCs) is establishment of food industries and related distributions in appropriate places. In this respect, geospatial technology offers the most up-to-date Land Cover (LC) information to improve site selection for assisting food security in the study area. Currently food security issues are not comprehensively addressed, especially in DSCs. In this research, ASTER L1B and LANDSAT satellite data were used to derive various LC biophysical parameters including build-up area, water body, forest, citrus, and rice fields in Qaemshahr city, Iran using different satellite-derived indices. A Product Level Fusion (PLF) approach was implemented to merge the outputs of the indices to prepare an improved LC map. The suitability of the proposed approach for LC mapping was evaluated in comparison with Support Vector Machine (SVM) and Artificial Neural Network (ANN) classification techniques. For implementing site selection, the outcomes of satellite-derived indices, as well as the city, village, road, railway, river, aqueduct, fault, casting, abattoir, cemetery, waste accumulation, wastewater treatment, educational centre, medical centre, military centre, asphalt factory, cement factory, and slope layers were obtained using Global Positioning System (GPS), on-screen digitizing, and image processing were used as input data. The Fuzzy Overlay and Weighted Linear Combination (WLC) methods were adopted to perform site selection process. The outcomes were then classified and analyzed based on the accessibility to main roads, cities and raw food materials. Finally, the existing industrial zones in the study area were evaluated for establishing food industries based on site selection results of this study. The results indicated higher performance of PLF method to provide up-to-date LC information with an overall accuracy and Kappa coefficient values of 95.95% and 0.95, respectively. The site selection result obtained using WLC method with the accuracy of 90% was superior, thus it was selected for further analyses. Based on the achieved results, the study has proven the applicability of current satellite data and geospatial technology for food industry site selection to resolve food security issues. In conclusion, site selection using geospatial technology provides a great potential for a reliable decision-making in food industry planning, as a significant issue in agro-based food security, especially in sanctioned countries.