GIS-Enabled Digital Twin System for Sustainable Evaluation of Carbon Emissions: A Case Study of Jeonju City, South Korea

Abstract

Despite the growing interest in digital twins (DTs) in geospatial technology, the scientific literature is still at the early stage, and concepts of DTs vary. In common perspectives, the primary goals of DTs are to reduce the uncertainty of the physical systems in real-world projects to reduce cost. Thus, this study is aimed at developing a structural schematic of a geographic information system (GIS)-enabled DT system and exploring geospatial technologies that can aid in deploying a DT system for a real-world project—in particular, for the sustainable evaluation of carbon emissions. The schematic includes three major phases: (1) data collection and visualization, (2) analytics, and (3) deployment. Three steps are designed to propose an optimal strategy to reduce carbon emissions in an urban area. In the analytics phase, mapping, machine learning algorithms, and spatial statistics are applied, mapping an ideal counterpart to physical assets. Furthermore, not only are GIS maps able to analyze geographic data that represent the counterparts of physical assets but can also display and analyze spatial relationships between physical assets. In the first step of the analytics phase, a GIS map spatially represented the most vulnerable area based on the values of carbon emissions computed according to the Intergovernmental Panel on Climate Change (IPCC) guidelines. Next, the radial basis function (RBF) kernel algorithm, a machine learning technique, was used to forecast spatial trends of carbon emissions. A backpropagation neural network (BPNN) was used to quantitatively determine which factor was the most influential among the four data sources: electricity, city gas, household waste, and vehicle. Then, a hot spot analysis was used to assess where high values of carbon emissions clustered in the study area. This study on the development of DTs contributes the following. First, with DTs, sustainable urban management systems will be improved and new insights developed more publicly. Ultimately, such improvements can reduce the failures of projects associated with urban planning and management. Second, the structural schematic proposed here is a data-driven approach; consequently, its outputs are more reliable and feasible. Ultimately, innovative approaches become available and services are transformed. Consequently, urban planners or policy makers can apply the system to scenario-based approaches.