Prototype Design of a Data Warehouse for Determining, Mapping, Monitoring and Visualizing Urban Heat Islands: The Case of Zagreb and Split, Croatia

Abstract. The UHI phenomenon is not visible to the naked eye and its intensity changes depending on the time of day and year. Natural and artificially induced phenomena such as the creation of Urban Heat Islands (UHI) are becoming more frequent and intense in recent years. Data warehouse combined with a decision support system (DSS) are offered as good solutions to this task. Research presented in this paper focuses on monitoring the phenomenon of UHI and providing guidelines for designing a specific data warehouse for that purpose. In this conceptual preliminary communication, theoretical considerations on the design of the future data warehouse for the detection and prevention of the creation of UHI, and decisions on the selection of individual parts in it, were presented. A data warehouse design concept using Online Analytical Processing (OLAP) type is presented with the star schema modelling technique, which will contain a catalogue of data needed to detect new and monitor existing UHIs in the cities of Split and Zagreb, Croatia. The single-tier model was chosen for the data warehouse architecture, due to its specificity and narrow scope, and the Extract, Load, Transform (ELT) tool will be used within it. A data warehouse system designed in this way, in combination with a decision support system (DSS) that should utilize its full potential, should ensure smooth data collection and efficient processing, and provide the possibility of transferring the methodology to other areas of big data processing.