[go: up one dir, main page]
Next Article in Journal
Mobility Prediction and Resource-Aware Client Selection for Federated Learning in IoT
Previous Article in Journal
Decentralized and Secure Blockchain Solution for Tamper-Proof Logging Events
Previous Article in Special Issue
User-Generated Content and Its Impact on Purchase Intent for Tourism Products: A Comparative Analysis of Millennials and Centennials on TikTok
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Future Internet
Volume 17
Issue 3
10.3390/fi17030110
futureinternet-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Developing an Urban Digital Twin for Environmental and Risk Assessment: A Case Study on Public Lighting and Hydrogeological Risk

by
Vincenzo Barrile
1,*,
Emanuela Genovese
1,
Clemente Maesano
2,
Sonia Calluso
2 and
Maurizio Pasquale Manti
1
1
Department of Civil Engineering, Energy, Environment and Materials (DICEAM), Mediterranea University of Reggio Calabria, Via Zehender, 89134 Reggio Calabria, Italy
2
Department of Civil, Building and Environmental Engineering, University of Rome “La Sapeinza”, Via Eudossiana, 18, 00184 Rome, Italy
*
Author to whom correspondence should be addressed.
Future Internet 2025, 17(3), 110; https://doi.org/10.3390/fi17030110
Submission received: 16 January 2025 / Revised: 17 February 2025 / Accepted: 21 February 2025 / Published: 1 March 2025
(This article belongs to the Special Issue Advances in Smart Environments and Digital Twin Technologies)
Browse Figures
Figure 1
<p>Research flow.</p> "> Figure 2
<p>Sensors’ acquisition system: (<b>a</b>) complete system for the acquisition of sensor data; (<b>b</b>) part of the sensors installed in the city center of Reggio Calabria.</p> "> Figure 3
<p>Urban digital twin architecture.</p> "> Figure 4
<p>Study area; Municipality of Reggio Calabria (Italy).</p> "> Figure 5
<p>Preview of the urban digital twin of Reggio Calabria: 3D representation of light pollution expressed in radiative flux.</p> "> Figure 6
<p>Preview of the urban digital twin of Reggio Calabria: 3D representation of hydrogeological risk.</p> "> Figure 7
<p>Correlation matrix between light pollution and CO<sub>2</sub>.</p> "> Figure 8
<p>Correlation matrix between land use, risk area, rainfall, and slope.</p> "> Figure 9
<p>Distribution of hazard areas by land use class.</p> ">
Versions Notes

Abstract

:
Improvements in immersive technology are opening up new opportunities for land management and urban planning, enabling the creation of detailed virtual models for examining and simulating real-world short-, medium-, and long-term scenarios. The goal of this research is to present the creation of an urban digital twin based on a virtual reality city replica, that models and visualizes the urban environment in three dimensions using advanced geomatics techniques and IoT technologies. The methodology focuses on two case studies that utilize environmental analysis and virtual simulation: assessing hydrogeological risk and evaluating public light pollution. The Cesium platform was employed to build high-precision 3D models based on topographic, meteorological, and infrastructure data. The proposed methodology calculated a correlation between light pollution and CO2 equal to 0.51 and a correlation between precipitation, slope, and risk area higher than 0.80. The most critical and high-risk classes are as follows: Dense Discontinuous Urban Fabric, Roads and Associated Lands, Pastures, and Forests. Results show how an urban digital twin can be a powerful tool for monitoring and territorial planning, with concrete applications in the public and risk management fields. This study also highlights the importance of geomatics technologies in the creation of realistic and functional virtual environments for the assessment and sustainable management of urban resources.

1. Introduction

Urbanization, with its rapid and continuous expansion, has brought significant transformations to cities, but it has also generated increasingly complex environmental [1], social, and economic [2] challenges. Issues such as light pollution [3,4], high energy consumption [5], and natural hazards such as hydrogeological risks emphasize the urgency of adopting innovative solutions based on accurate data and multidimensional analysis. In detail, [6] provides a review of the scientific literature on light pollution, while [7] proposed a multimodal machine learning model called LPS-Net (Light Pollution Assessment Network Model) to assess light pollution by considering various indicators such as ambient lighting, light color, and exposure time to the light source. Regarding the hydrogeological hazard, [8] proposed a review of flood modeling in urban developments, highlighting the various strengths and weaknesses and emphasizing the importance of model calibration to adapt them to local situations and case studies. The paper also provides valuable insights into understanding the evolution of flood hazards at different scales.
Emerging technologies, such as 3D platforms and digital twins, have proven to be extremely effective tools for representing, analyzing, and simulating complex scenarios [9,10,11], offering a visually intuitive and technically advanced approach to addressing challenges related to sustainability and urban resilience. For example, researchers in [12] proposed a 3D platform able to visualize energy data of buildings; another study [13] proposed a 3D platform, enhanced with AI, to provide improved damage evaluation and modeling. In the latter, Ref. [14] explores the role of digital platforms in providing faster responses and improving urban resilience.
In this context, the urban digital twin enables the integration of geomatic [15,16,17], infrastructural, and meteorological data into a single platform, supporting strategic decision-making processes. The ability to create realistic simulations and predictive analyses promotes the adoption of more informed and efficient land management policies. The concept of a digital twin is relatively new and still in its early stages, but extensive research and practical applications are gaining momentum in all fields of research, from the creation of digital twins for industries and plants, to cars and airplanes, to road structures and infrastructure, and even to healthcare facilities (hospitals and outpatient clinics).
In this context, this research work aims to demonstrate the applicability and potential of the urban digital twin in sustainable territory management and environmental risk assessment, overcoming some of the characteristic limitations of its implementation. Specifically, in the following:
  • Public lighting management, aimed at reducing light pollution and improving energy efficiency.
  • Hydrogeological risk assessment, with simulations of flood events to identify vulnerable areas.
To achieve this goal, various data sources were used to characterize the studied phenomena (lighting and hydrogeological risk), acquired through different Geomatics and Earth Observation techniques, including satellite data, drone data, GNSS measurements, and ground-truth sensors. For the visualization and management of the collected data, a system based on Cesium technology was implemented, enabling the operation of three-dimensional and geospatial data within a single, interactive environment.
The development of this integrated and updatable system provided the current state of the chosen study area, a portion of the metropolitan city of Reggio Calabria. The digital twin was developed in an open-source environment with open data, allowing users to upload their own data to the platform and use it with custom-developed models. Thus, the first experimental results of this research are presented, demonstrating how a highly interactive digital environment can serve as a valuable tool for urban planning and risk management, contributing to the sustainability and environmental resilience of cities.
This paper is structured as follows: the first section is dedicated to the introduction, where the framework and the main theme of the proposed study are presented. A subsection is dedicated to a detailed analysis of the main technologies already present at national and international levels, highlighting the challenges encountered and the innovation of the proposed approach. Section 2 focuses on materials and methods, where the IT procedures and the methodology for the realization of the urban digital twin are detailed. Section 3 is dedicated to the qualitative and quantitative presentation of the obtained results. In Section 4 the main limitations and future developments of the proposed technology are discussed. Finally, Section 5 is dedicated to the conclusions.

Related Works

In the literature, several authors have focused on the implementation of such technology, which should represent the dynamic and virtual replica of a structure or process. For example, Ref. [18] proposes a digital twin of the AANET system, which consists of aviation networks used to solve the problem of connectivity in dynamic environments. The authors present an integrated and complex system that includes the implementation of the virtual physical environment, the operational module, and the use of supervised and unsupervised learning techniques, typical of artificial intelligence, to optimize the network.
Most notable is the work of [19], in which the creation of a digital twin model to test the operation of an electric car is presented, using AC motor current as input data and DC motor rotational speed as output. It also includes satellite data for terrain modeling and analysis software for validating the results with historical data. Another significant contribution in the field of digital twin applications in the industrial sector is that of [20], which introduces a digital twin for the automotive industry, focusing on the continuous optimization of production processes, proactive maintenance, and the ongoing processing of process data.
In the area of transportation and infrastructure, the contribution of [21] stands out, highlighting the need for more efficient infrastructure systems, achievable through the implementation of cyber–physical systems and, therefore, an intelligent digital twin. A case study of a bridge is presented, analyzing in detail the steps necessary to implement the system and providing an overview of the physical, cyber, and integration details. The agricultural sector is also increasingly exploring the development of digital twins [22,23]; in detail, Ref. [24] has developed a Decision Support System using a digital twin related to aquaponic cultivation.
Even more interesting is the application of such technology in the field of healthcare, specifically regarding the implementation of digital twins of hospital facilities (hospitals and outpatient clinics) and, once again, the possibility of creating one for the human body. This is extensively explored by [25], which compares the traditional digital twin and the human digital twin, highlighting the potential for realization through new technological advancements and proposing a system architecture and implementation approach. In healthcare, Ref. [26] has developed a pragmatic example of a smart ward at Shanghai Tongji Hospital.
While the construction of a digital twin of a process, facility, or technology can be governed by knowledge of the structure of the object or system in its process, the digital twin of a city and territory (being a natural process comparable to the complex processes of the human body) is undoubtedly a more intricate and dynamic endeavor. It must take into account several variables that change continuously, even significantly altering the source information. In this context, several authors have shown interest in the implementation of this technology [27,28,29]. In particular, the authors in [30] focused on contributing to the conceptualization of the urban digital twin. They mapped the initiatives in terms of applications, inputs, processes, and outputs, identifying the requirements for properly structuring this emerging technology. A comprehensive review of these challenges is presented by [31], which outlines the technical challenges related to implementing urban digital twins, including data quality and standards, interoperability, data integration, complexity, and data ownership.
In this field, worthy of citation is the platform Snap4City. Snap4City (powered by Fiware) is a relevant platform developed by the University of Turin (also in Italy) for multiple purposes [32]. It is a solution that provides users with access to various services (from traffic to air pollutants) and allows the addition of new services through personalized access after registration. The platform enables the creation of dashboards for a range of use cases, from city management to industry. The technologies used are of high value and are cutting-edge, and the platform is presented as versatile and scalable, capable of adapting to various desired solutions. It also includes personalized videos and tutorials, as well as numerous informative documents on its usage, making the user’s experience easier.
Regarding the integration of digital twins in urban planning, worthy of note is also the case of the city of Zurich [33]. In detail, the paper analyzed the digital twin of the city of Zurich which is composed of high detailed 3D and GIS Data and is suitable for several urban planning-related applications such as the development of the Municipal Development Plan, for planning the high-rise building, heat reduction, and spread of the urban planning procedure across the population. Another example is provided in [34], where the authors presented the digital twin of the city of Herrenberg, built using a combination of 3D modelling, graph theory applied for street network modelling, wind and urban mobility simulation, and a sensor data network. This model was also validated through a questionnaire distributed to 39 participants. For what concerns the use of digital twin technology for environmental risk management, Ref. [35] proposes an innovative digital twin related to water management, climate change adaption, and disaster risk reduction, introducing real-time dynamic updating into the DK-HIP (hydrological information and prediction) model. The paper also explores the potential for developing HIP digital twins for local rivers and basins. Additionally, Ref. [36] discusses the scientific concept of intelligent disaster prevention and mitigation for infrastructure, based on digital twin technologies. The authors of [37] proposed an innovative framework regarding urban flooding digital twin development; the framework is based on a user-centered design process and includes two main components: a UFDT conceptual model and a generative methodology for its rapid construction and update. Lastly, an in-depth review of the evolution of digital twin smart cities regarding disaster risk management is provided by [38] and by [39]. In recent years, the development of the urban digital twin has seen several improvements, particularly in what concerns risk management. One example is presented in [40] where the authors proposed a digital twin as a solution to address urban heat island challenges, such as higher energy consumption and lower air quality, also offering a case study in the city of Enschede in the Netherlands. Another study [41] developed an urban digital twin with a geospatial dashboard to enable visualization of critical infrastructure vulnerabilities across a range of spatial and temporal scales.
In this framework, the main technological limitations are interoperability, multi-scale modeling and simulation, scalability and data management, as well as in their applications in emerging sectors. Our research focuses on optimizing real-time updates through a distributed architecture based on an IoT sensor network, ensuring continuous and high-frequency data acquisition. The adopted multi-scale modeling allows the integration of heterogeneous data from different sources, providing a high-fidelity representation of the physical system. An additional challenge in DTs is the management of large volumes of data generated in real time. In our approach, this problem is mitigated using edge computing techniques, which allow local data processing directly at the acquisition sources, reducing latency and optimizing the transfer of information to the cloud or other computational nodes. Furthermore, the proposed framework addresses scalability issues thanks to an optimized architecture for the management of fluid visualizations even on high-resolution geospatial datasets, preserving the integrity and reactivity of the system. The application of the developed methodology is particularly relevant for the analysis and management of urbanized environments, an emerging sector that requires advanced monitoring and forecasting tools. Our work has been implemented in a geographical context of particular interest, the Calabria Region (Italy), characterized by critical issues related to the management of hydrogeological risk and the optimization of urban lighting. The integration of the digital twin in this scenario allows for the support of risk mitigation and energy efficiency strategies, providing an innovative operational model for territorial governance based on dynamic and real-time updated data.

2. Materials and Methods

The design of the urban digital twin was based on an advanced digital architecture, focused on creating an interactive virtual environment to support environmental analysis, strategic planning, and eventually accommodating forecast simulation models. The system is fully open, allowing the uploading of pre-processed data for customized forecasts. The construction of the system was articulated in several stages, starting with the selection of the most suitable technological platform and ending with the integration of data from different, highly specialized sources. For this experimental research, the Cesium Ion platform was chosen—an open-source platform that allows visualization, management, and analysis of 3D geospatial data. The following diagram (Figure 1) functionally summarizes the various steps followed in the design and development of the proposed experimental system:
The process described is divided into several phases: The first phase involves the collection and cataloging of data useful for the planned simulations. The data collected for characterizing a process or a physical phenomenon is often of various types and comes from different sources. Therefore, efficient collection and cataloging are necessary to avoid redudancy or missing data. The second phase is mainly dedicated to the processing and integration of this heterogeneous data, which must be communicated within a single interoperable system. The third phase is focused on the creation of a physical model that not only represents the data but also allows their interaction, allowing them to evolve based on the results that need to be achieved. This allows progression to phase four, the management phase, in which it is possible to analyze the current state of the phenomena that characterize the digital twin. Phases five and six are closely interconnected. Starting from the mapped and analyzed current state, they aim to create future or hypothetical scenarios and projections that can support proactive territory management, facilitating decisions based on data and forecasts.
In order to apply the proposed technology, two phenomena closely linked to industrialization, global warming, and climate change were evaluated: light pollution and hydrogeological risk.
In relation to public lighting, the aim of this work was to conduct advanced simulations to analyze different aspects of the public lighting network, specifically examining the influence of lighting on the concentration of CO2 in a densely populated urban center and their close correlation. In relation to hydrogeological risk, the aim was to evaluate the effect of climate change on flood-related risks. The results show the experimental methodology and propose the creation of representative models of such phenomena, which still require in-depth studies.

Implementation of the Proposed Methodology

The first phase involved the collection of heterogeneous data from different data sources. In particular, we gathered static data for mapping the territory (data from drones and satellites) and dynamic data for real-time monitoring (data from sensors and meteorological stations).
A technique for mapping the territory on a large scale is certainly remote sensing, which includes both satellite images and drone images. For this research work, different types of satellite images were acquired from various services available free of charge.
In relation to the phenomenon of light pollution, data from the VIIRS (Visible Infrared Imaging Radiometer Suite) sensor was collected, visualized, and analyzed [42]. The sensor is equipped with 22 spectral bands, ranging from visible wavelengths (0.4–0.9 μm) to thermal infrared (10.3–12.5 μm), covering the visible spectrum, SWIR, TIR, and microwave. Mounted on satellites in polar orbit, the sensor provides detailed measurements of artificial lights during the night and offers sufficient resolution (up to about 750 m per pixel) to detect and monitor variations in light pollution in relation to the thermal bands, while the resolution is 1.5 km per pixel. The data are distributed by portals such as NOAA CLASS, NASA Earthdata, and VIIRS Nighttime Lights and provide global daily coverage thanks to the use of the Day/Night Band (DNB) channel, which is particularly sensitive to light even in low-light conditions, such as during the night phases. Several images were then acquired from this dataset over the study area to analyze changes in artificial brightness (expressed in radiative flux) and identify the most sensitive areas in relation to this portion of the territory. Alongside this dataset, data from the Copernicus Earth Observation service were also acquired, offering a wide range of satellite images and atmospheric variables. Specifically, data on carbon dioxide was obtained through the CAMS service (Copernicus Atmosphere Monitoring Service), including the CAMS global emission inventories [43]. This service also provides accurate estimates for other greenhouse gases, such as methane, nitrogen oxide, and other polluting compounds. It also offers both concentration distribution estimates and flux data (emissions and natural absorption).
Regarding dynamic data, sensor data were used. First, an accurate and detailed mapping of the light sources was carried out by conducting a census with GNSS Instruments (Rover GNSS Sanding T5). For each light point, the precise latitude and longitude coordinates were recorded, ensuring accurate georeferencing. Each light source was cataloged based on the type of lamp used (e.g., LED or traditional), the power expressed in watts, the color temperature (Kelvin), and the maintenance status to identify any critical issues. The aim was to conduct an in-depth comparison of the performance of LED lamps, recognized for their high efficiency and sustainability, while identifying possible areas for system optimization and intervention.
Regarding the phenomenon of hydrogeological risk, the data were acquired from meteorological stations situated in various areas. The annual datasets provide information on precipitation and flow data from the stations distributed across the metropolitan city, which were processed to make them compatible with the proposed system. Additionally, data in GeoJSON format of buildings and watercourses were retrieved.
After the first phase, which focused on the collection of data essential for characterizing the phenomena of light pollution and hydrogeological risk, we proceeded to the second phase, which involved the processing and integration of this data into a structured database. This database can be visualized within a 3D geospatial data visualization system. For the management and visualization of geospatial data, it was decided to use the Cesium library, which was conveniently integrated into an HTML file running on a local server. Cesium is a powerful JavaScript library that allows the creation of interactive 3D visualizations on a globe, ideal for visualizing data from sensors or GIS systems. To efficiently manage geospatial and non-geospatial data, PostgreSQL with the PostGIS extension enabled was chosen. PostGIS extends the functionality of PostgreSQL by adding support for spatial objects, allowing complex operations such as geospatial analysis, geocoding, and spatial queries. Vector data (for example, polygons and lines), raster data (satellite images), and data from sensors were inserted into the geodatabase (Figure 2). To ensure smooth interaction with the database and enable querying capabilities, a backend server was implemented. This server exposes a RESTful API for communication between the frontend (Cesium) and the database. It allows for the real-time retrieval of geospatial data, performs analytical operations, and displays the results directly on the map.
To have an accurate three-dimensional representation of the study area, a flight was carried out with a DJI Matrice 350 RTK drone (DJI, Shenzhen, China) equipped with a LiDAR sensor. This UAV features a diagonal wheelbase of 895 mm and a weight, without batteries, of about 3.77 kg. The operating frequencies are between 2.4000–2.4835 GHz, 5.150–5.250 GHz, and 5.725–5.850 GHz. The maximum horizontal speed is 23 m/s, and the maximum flight autonomy is 55 min [44].
After the flight, the point cloud was pre-processed within the CloudCompare software (version 2.13.2), and the 3D model was created within the Metashape software (version 2.2.0). Once the modeling part was completed, the model, exported in .fbx format, was converted into .gltf format to ensure its compatibility with Cesium. The conversion process included transforming the mesh into a format compatible with WebGL, managing the texture maps, and translating the materials used in the model into a scheme that Cesium could interpret correctly. The GLTF format was chosen because it is designed to be lightweight and easily downloadable from the web. It is more compact than other formats, such as FBX, and is suitable for contexts that require real-time streaming of 3D models, such as web applications and augmented and virtual reality platforms, which are the foundation for the creation of the digital twin and the metaverse. GLTF is closely linked to WebGL technology, and being a JSON-based format, it is natively compatible with Javascript libraries such as CesiumJS and Three.js, which are used in 3D rendering platforms based on WebGL. Furthermore, for correct data integration within web visualization, it supports a standard representation of materials through the PBR (Physically Based Rendering) standard. Obviously, the 3D model was georeferenced, i.e., aligned to the correct geospatial coordinate system (EPSG:4326 WGS84), so it could be correctly positioned on the surface representing the study area within Cesium. Table 1 is a summary of the data source used in this work.
The different geospatial data collected in the first phase were then superimposed on this 3D model. In particular, for the light pollution analysis, data collected using GPS instrumentation were processed, followed by a clustering operation based on the parameters of diffuse radiation emitted by the lighting devices. For this purpose, the DBSCAN algorithm was used, which enabled the identification of areas with the highest levels of light pollution during nighttime hours. This approach helped pinpoint the regions that could have the most significant impact on CO2 concentration.
For the hydrogeological risk analysis, data on flow rate, elevation, and land cover for the study area were processed. These data served as input to the 2D hydraulic model used for simulation, capable of reproducing the dynamics of water flow: HEC-RAS (v. 6.6). This model uses the Saint-Venant equations to simulate the water flow in canals and rivers and incorporates different relationships related to mass and momentum, namely the Continuity Equation, Momentum Equation, and Manning’s formula. In HEC-RAS, the diffuse wave approximation of the Saint-Venant equations is used to simplify the calculation in 2D simulations. The equations are solved in two dimensions, considering both the variation of the flow along the watercourse and transversally.
To define the flood area, the DEM (Digital Elevation Model) of the study area to be analyzed along with GeoJSON files of the watercourse affecting the area and the buildings were used. The entry and exit conditions were then set using data from the meteorological stations, particularly flow data. In this regard, Earth Observation data provided valuable and detailed information about the study area. Although direct validation with historical data has not been conducted at this stage, the HEC-RAS model is widely recognized in the scientific community for its ability to simulate flood phenomena with high accuracy. In addition, the consistency of the results obtained with the morphology and hydrographic network of the study area confirms the robustness of the simulations.
In a digital twin, it is not enough to have only a static representation of the physical world (in this case, the configuration of the city’s urban layout). For this reason, data relating to the concentration of CO2 were collected from the sensors. These data were acquired using a microcontroller, the ESP32, equipped with Wi-Fi and Bluetooth connections (Figure 2) and a Sen63C sensor (Sensirion), which measures pollutants such as temperature, relative humidity, atmospheric particulate matter, and carbon dioxide at different points in the city, taking two measurements per day. Figure 2 shows the microcontroller, which has been integrated with a CO2 concentration sensor, a central component of the technology used in this study, which, thanks to a Wi-Fi module, sends the measured data to the server via a network connection. Once received, the data are stored in the centralized database, using a data management system that can guarantee scalability and integrity of the information. The backend then processes the data, enabling the visualization of results through a WebGIS interface that facilitates interactive, real-time use of geospatial information. Figure 2a shows an example of a sensor installed in a city context, highlighting its position in the urban ecosystem, while Figure 2b presents a map that displays the distribution of sensors during the monitoring period, allowing observation of the coverage of the territory for real-time data collection.
Communication between the microcontroller and the platform is in this way ensured using backend software. Through the appropriate libraries, the data can be displayed through graphs, dashboards, and other visualizations.
Data preprocessing, cleaning, and integration, as known, are fundamental steps to ensure effective use of data and to make the analyses both consistent and robust. Regarding data preprocessing, the first phase consists of data collection and aggregation. The acquired satellite data were in GeoTIFF format, the drone data were in LAS format, while the data collected by sensors were provided in JSON format. Preprocessing included temporal alignment (through a data interpolation process to align them to a common time interval) and geospatial alignment, projecting all data into a common coordinate reference system (WGS84). For some of the data, spatial resolution alignment was necessary to standardize the resolution across all datasets. Regarding data cleaning, this phase involved the process of removing anomalous data and outliers from sensor data. The integration process involved the rasterization of sensor data to enable analysis alongside satellite data and the vectorizing data obtained from hydrogeological risk modeling software. The integration also included geospatial overlay through data loading from databases and the use of appropriate scripts in the frontend for visualization. Clean and integrated data were stored in spatial databases, while the visualization was generated using the Cesium JS viewer (version 1.87).
The prototype and experimental system implemented must also simulate physical dynamics, such as changes in the elements that constitute the virtual environment, including the deterioration of infrastructure, changes in vegetative cover, and other evolving phenomena. In this context, Cesium enables the application of animations to models representing these changes over time. This can simulate various dynamic processes, such as vehicle movement, pedestrian flow, urban growth, and the evolution of climate conditions [50].
Furthermore, Cesium can integrate advanced spatial analysis tools to calculate, for example, the travel time between two points based on traffic, evaluate the impact of certain extreme events on the territory, and create simulations of the physical world’s changes in user-defined scenarios. These capabilities are supported by algorithms designed to interpret the inherent characteristics and natural dynamics of the phenomena being analyzed.
Subsequently, in Figure 3, the IT structure of the urban digital twin developed is presented.
The database, as previously mentioned, was loaded using the PostgreSQL (version 17.4) application, a powerful open-source relational database management system that offers a wide range of advanced features for data management and processing. For the specific needs of this project, the PostGIS (version 3.5.0) extension was also enabled. This PostgreSQL extension allows for the management of geospatial data, that is, data containing information about a geographic position. Thanks to PostGIS, it is possible not only to store georeferenced data (such as geographic coordinates), but also to perform complex spatial operations, such as calculating distances, searching for geometries within certain areas, and manipulating geographic shapes like polygons and lines.
Using PostgreSQL with the PostGIS extension offers significant advantages over other geospatial data management systems. Its ability to manage large volumes of data and its reliability make it an ideal choice for projects that require a solid foundation for archiving and processing complex data. In addition, PostgreSQL allows for different modes of interaction: data can be loaded and managed via its graphical interface, which is useful for less experienced users, or via the command line using the psql application, offering greater flexibility for more experienced users who want to manage the database with custom queries.
The code that manages the visualization and interaction with the urban digital twin is contained within an HTML file. Once loaded into the browser, this file allows access to the application. In addition to the basic structures of web programming, the HTML file includes fundamental components for visualizing and interacting with geospatial data. In particular, one of the main libraries used is Cesium. With Cesium, the system can visualize the geospatial data loaded into the platform and enable navigation through it, making the user experience more engaging and dynamic. Cesium also allows for the visualization of objects and urban structures and provides advanced visualization operations, such as rendering 3D models.
In addition to Cesium, the HTML file includes other specialized JavaScript libraries that enable advanced spatial queries. These libraries are essential for managing communication between the database and the frontend, as well as providing additional features like geometry analysis, elevation management, and temporal data visualization. These features enable the simulation of various urban scenarios over time, such as the evolution of an area or changes in environmental conditions. The integration of these libraries enriches the user experience and ensures effective spatial analysis, which is crucial for monitoring and managing complex geospatial data.
The interactivity of the platform is ensured by the two-way communication between the frontend and the database, facilitated via a backend server. In this case, the backend server is managed by a Node.js script, a JavaScript runtime environment that allows the execution of JavaScript code on the server. Node.js is particularly suitable for applications requiring high performance and scalability, such as those handling large amounts of data in real time. The Node.js server acts as an intermediary between the frontend, which interacts with the user, and the backend, which manages the database.
This architecture allows data to be loaded at startup through a REST API connection, a methodology that ensures secure and efficient data exchange between the client (browser) and the server. By using a REST API, the system avoids loading the entire dataset into memory at startup, reducing loading times and improving the fluency of the browsing experience. In practice, only the data needed for the initial visualization are loaded, while the remaining data are dynamically retrieved via API requests when the user interacts with the platform. This significantly lightens the local server and optimizes overall performance.
Furthermore, communication between the frontend and the backend ensures that all changes made by the user—both in the visualization interface and in spatial analysis operations—are synchronized in real time with the database. This means that any update or change made to the platform, such as adding new data or processing new spatial analyses, is immediately reflected in the database, keeping the data always up to date and consistent. This is particularly useful in contexts where the platform is used by multiple users simultaneously, allowing them to collaborate and share results in real time.
As far as satellite data, raster, vectors and drone flight data are concerned, these have been stored in a cloud solution within a geospatial database, based on a scalable and high-performance platform. This is a flexible, manageable solution whose limit is the storage space, which is currently around 10 TB, thus widely supporting the saving of the data analyzed so far. For the storage of these large volumes of data, a geospatial data compression system has been implemented using optimized formats such as GeoTIFF with lossless compression, reducing the volume of stored data and improving efficiency in reading and writing operations. Furthermore, the data were organized in a PostGIS database, which allows for efficient management of spatial data with advanced geospatial query operations, reducing response times and optimizing data access.
To process data from the sensors more efficiently, an edge computing process was implemented that allowed the data to be processed locally, close to the acquisition source, instead of sending it directly to the server. For this purpose, advanced data filtering methods were used to eliminate outliers through standard deviation analysis. Furthermore, a data aggregation algorithm was applied on pre-established time intervals, using a moving window, to further reduce the volume of data to be transmitted and to ensure a more stable and precise view of the environmental variables. The aggregation process was implemented on high-efficiency devices, such as ESP32 microcontrollers, which perform real-time filtering and data aggregation operations in a distributed manner.
The advantages of this operation include the reduction in communication volume since not all unnecessary data are transmitted. This approach reduces network traffic and allows only relevant information to be transmitted. From an energy efficiency perspective, processing data locally reduces the power consumption of the device. ESP32 microcontrollers used for edge processing require less energy to perform local computation operations, compared to transmitting large volumes of raw data to a central server. In addition, implementing real-time compression algorithms on edge devices further reduces the volume of data sent, improving network efficiency.
From a communications network perspective, lightweight communication protocols such as MQTT (Message Queuing Telemetry Transport) have been implemented to optimize bandwidth usage, suitable for IoT applications and capable of efficiently managing messages between devices and servers. MQTT allows optimized network management, reducing the need to constantly transmit large amounts of data, thanks to its pub/sub (publish/subscribe)-based architecture, which sends only the necessary data at the right time.

3. Results

3.1. Study Area

The study area, selected for this experimental research and the implementation of the urban digital twin is the city center of Reggio Calabria (Figure 4). This city has a unique geographical conformation, where urban development is primarily concentrated along the coastal area, gradually expanding toward the mountains, and eventually meeting hilly and peripheral zones. This layout makes Reggio Calabria an ideal location for studying both light pollution and hydrogeological risk phenomena.
The contrast between the city center and its outskirts allows for a clear distinction in light pollution levels, providing valuable insights into the impact of urbanization on artificial light distribution. Additionally, the city’s location makes it particularly suitable for analyzing hydrogeological risks. Reggio Calabria is home to natural watercourses, such as the “fiumare” typical of Mediterranean regions, which are prone to flooding, especially during periods of heavy rainfall.
According to the ISPRA (Higher Institute for Environmental Protection and Research), the hydrological risk, including flooding and landslides, is increasing, particularly in the Calabria region [51]. The city’s susceptibility to these phenomena makes it an important case study for understanding the potential impacts of climate change and urbanization on flood risk and environmental resilience. The image in Figure 5 illustrates the study area chosen for this research.

3.2. Illumination

In Reggio Calabria, public lighting is not fully standardized. While there is a growing trend toward adopting LED lamps, many of the city’s streets are still illuminated using older technologies, such as high-pressure sodium lamps. These lamps have been standard in many European cities for several years. Given the environmental and economic benefits of LED lamps—including greater energy efficiency, longer lifespan, and improved control over light direction and dispersion—it is particularly beneficial to introduce LED lighting in the city.
To optimize public lighting and manage light pollution effectively, an accurate mapping of the current lighting situation is essential. Advanced technological tools, such as digital simulations, can help identify the best interventions for regulating lighting. By using satellite data on light pollution and the actual specifications of the streetlamps in the city, we conducted an initial analysis of light pollution levels within Reggio Calabria.
Additionally, data obtained from on-site measurements were also analyzed and integrated into the system. Figure 5 presents the results from this proposed methodology, highlighting the initial findings related to the city’s light pollution and the potential impact of different lighting technologies.
The representation shows the intensity of light pollution, which increases with distance from the city center but also reveals significant issues in the suburbs. The initial results obtained from the proposed system can be continuously updated by incorporating a mathematical model. Once defined, this model can be integrated into the platform, which is already structured to accommodate such functional models. The model will adjust the results in real-time based on input variables from brightness sensors deployed across the territory and the acquired satellite images.
This system not only enables the real-time visualization of changes but also allows for the simulation of modifications, analyzing their potential impacts. The goal of this model is to facilitate evaluations based on different territorial scenarios and, potentially, to monitor the evolution of light pollution over time. This ongoing development will further enhance the system’s ability to manage and mitigate light pollution effectively.

3.3. Hydrogeological Risk

The first results obtained have marked a significant advancement in the field of advanced digital modeling. The three-dimensional modeling provided a precise representation of the morphological characteristics of the territory, enabling the integration of accurate geospatial data. This integration allowed the identification of areas at risk based on land use, which is crucial for assessing and mitigating potential hazards.
The model was specifically designed to simulate hydrogeological risk scenarios, allowing for the evaluation of potential impacts on infrastructures, buildings, and the population. In this case, it was possible to identify the areas that are most susceptible to flooding within the city center, which is especially valuable for urban planning and emergency response.
By extending this study to incorporate various potential scenarios and configurations, the model can offer crucial insights into flood risk and related landslide risks. Such detailed assessments will support decision making in urban resilience and sustainability efforts, helping to develop proactive strategies for managing and mitigating environmental risks. Figure 6 shows in red the 3D representation of the urban center subjected to hydrogeological risk and in blue the basin area with the calculation of the stream segments of flow directions.
Overall, the proposed urban digital twin system architecture offers a multi-layered and flexible framework that supports the integration of real-time data from both on-site and satellite sensors. These sensors feed data into the system using the MQTT communication protocol, a lightweight and efficient solution for transmitting real-time data with low bandwidth consumption, making it ideal for IoT applications where frequent updates are required. Once the data are transmitted, they are processed and stored in relational databases, which simplifies data integration from various sources, ensuring that the system remains adaptable and scalable. This architecture allows the model to be easily extended to include new devices and additional data sources as they become available, ensuring its long-term relevance and versatility.
The simulation and experimental setup were designed to evaluate the accuracy and effectiveness of the urban digital twin in modeling flood risk and light pollution. For the flood risk simulation, the HEC-RAS (v. 6.6) 2D hydraulic model was employed to simulate water flow dynamics based on the Saint-Venant equations. Input data included high-resolution Digital Elevation Models (DEMs), land cover classification, river discharge data, and precipitation records. Boundary conditions were defined using historical flood events, and sensitivity analysis was conducted to optimize model parameters such as the Manning’s roughness coefficient. For the light pollution analysis, satellite data from VIIRS-DNB were integrated with ground-based light intensity measurements.
Regarding light pollution, the proposed methodology allowed the evaluation of the correlation between light pollution and the concentration of carbon dioxide in the study area. Correlation is shown in Figure 7. The analysis revealed a correlation of over 0.5 between these two variables, indicating a sufficient positive correlation. This suggests that as one variable increases, the other does as well.
Regarding the areas most affected by the phenomenon of light pollution, it was possible to estimate the results reported in the following Table 2:
The light pollution data detected by satellites and sensors were cross-referenced and integrated with land use data to quantitatively determine the areas that are most at risk of higher radiance during the night. The results reveal that the port area, along with the roads and railway, exhibits the highest concentration of illumination, in addition to the densely populated areas.
Regarding the hydrogeological risk, the correlation between slope, rainfall, land use, and the percentage of area with high hydrogeological risk was also calculated, as shown Figure 8. The results, shown in the figure, allow us to confirm a strong correlation (as expected) between rainfall and slope with the area at high hydrogeological risk, at 0.83 and 0.80, respectively.
We then proceeded to analyze the risk distribution through boxplots (Figure 9), which allowed us to highlight how some land use classes have a more concentrated distribution with few outliers, while others, such as the class “12220”, show a greater dispersion with extremely high values. Furthermore, several land use classes have points outside the boxplots, indicating the presence of anomalous values that could represent exceptional events or specific situations. Some categories, such as “12220” and “22000”, show a high variance and higher maximum values, suggesting that these types of land use could be more subject to risk phenomena. Finally, the classes with very compact boxplots and low values indicate that the risk area for those types of land use is generally limited. In summary
  • The most critical classes (12220, 22000, and 31000) have a wide dispersion and extremely high values, indicating a high risk.
  • The moderate risk classes (12230, 12300, 20000, 30000, 32000, and 33000) have a significant but not extreme variance.
  • The low-risk classes (11100, 11210, 11220, 11230, 13100, 13400, 14100, and 14200) show a narrower distribution with generally low values.
Similarly, for hydrogeological risk, we crossed the areas at high hydrogeological risk with the land use classes; results are shown in Table 3. The analysis revealed that the classes most at risk include Continuous Urban Fabric, Discontinuous Dense Urban Fabric, Discontinuous Medium Density Urban Fabric, Roads and Associated Land, and Port Area (within the city). Also of significance are the areas classified as Pastures and Herbaceous Vegetation Associations, which show notable vulnerability.
Light pollution has several adverse effects on public health and significantly impacts urban planning. The excess of artificial light is correlated with an increased risk of metabolic and cardiovascular disorders and also negatively affects the life of the city’s fauna. Several techniques can be adopted by public administration to reduce these effects, such as the use of LED lights, smart lighting systems, and the construction of ‘dark zones’ to preserve local fauna. Furthermore, although there is no direct relationship between light pollution and the concentration of carbon dioxide, this type of pollution is associated with increased energy consumption, mainly due to the higher energy use caused by unshielded external lighting, which contributes to CO2 emissions. In parallel, flood risk mitigation requires an integrated approach that involves the use of green infrastructure. The adoption of green roofs, permeable pavements, and the redevelopment of river areas can improve water absorption and reduce surface runoff. The use of early alert systems and predictive models can also support risk management, allowing rapid and effective responses to extreme events. Integrating these aspects into urban planning promotes cities that are more resilient, sustainable, and safe for people.
From a performance perspective, the system has shown promising results. CPU and memory usage have remained below optimal thresholds, ensuring efficient resource utilization. These results highlight the computational efficiency and effectiveness of the system, allowing for smooth management of large-scale urban data while maintaining high performance. This system architecture is highly scalable and provides the necessary infrastructure for future enhancements, making it well suited for urban planning, risk management, and environmental monitoring applications.
Compared to Barcellona’s traffic digital twin [27], Singapore’s smart cities initiatives [52], and Snap4City, our approach focuses on two specific phenomena, aiming to study and analyze their characteristics and evolution over time: light pollution and hydrogeological risk. The key difference in our approach is that users will not need to download any program or install anything on their PC and have access to the software as a service in cloud. With the data available, they will be able to work in cloud on solutions that go beyond viewing the uploaded data, enabling the creation of complex pipelines in which the results of one step’s processing become the inputs for subsequent steps. In detail, regarding Barcellona’s digital twin, this one is based on traffic simulation, aided by SUMO software (version 1.22) [51], and developed on a large scale; meanwhile, our digital twin is based on the Cesium library, and there also no software exception to local server, web browser, and database software. Our approach represents a less-computational resources solution also based on open-source software and libraries; the Singapore digital twin is mainly based on IoT and city sustainability while our project is mainly focused on hydrogeological risk and light pollution.
The aim of this work was also to connect the incoming data so that inputs can modify the proposed urban digital twin through relationships designed and studied specifically for this purpose. In fact, models characterizing the two phenomena will be implemented so that the digital twin is not just a data viewer but an integrated system capable of modifying the graphical representation based on input variables. This is made possible thanks to the integration of an IoT system that provides real-time data. Moreover, what differentiates our solution from the others is the 3D–4D visualization, which allows for an interactive and advanced visualization, even over time, of the phenomena studied via the bar at the bottom of the screen of the Cesium viewer.

4. Discussion

The urban digital twin we proposed stands out in several ways. First, our approach focuses on two specific phenomena, aiming to study and analyze their characteristics and evolution over time: light pollution and hydrogeological risk. The key difference in our approach is that users will not need to download any program or install anything on their PC. With the data available, they will be able to work in the cloud on solutions that go beyond viewing the uploaded data, enabling the creation of complex pipelines in which the results of one step’s processing become the inputs for subsequent steps. The aim is also to connect the incoming data so that inputs can modify the proposed urban digital twin through relationships designed and studied specifically for this purpose. In fact, models characterizing the two phenomena will be implemented so that the digital twin is not just a data viewer but an integrated system capable of modifying the graphical representation based on input variables. This is made possible thanks to the integration of an IoT system that provides real-time data.
To evaluate a digital twin, several aspects must be taken into account, both from the point of view of the obtainable results and from that of the system performance. From the point of view of the expected results, the predictions of the digital twin can be compared with the real data coming from the physical system that it replicates, calculating common metrics such as the RMSE or determining the accuracy of the predictions, that is, the system’s ability to correctly predict future events.
From the point of view of performance and scalability, there are several methods to evaluate the performance of the digital twin, which mainly focus on the efficiency in simulating the physical system, on its scalability in managing large amounts of data, and on the reliability in processing the data. In this sense, the proposed system works in the cloud, significantly reducing the computational load compared to local solutions. Furthermore, the system is easily integrated and interoperable with other data, allowing it to interact correctly with other systems and technologies. In the future, evaluation metrics may also include user experience and performance monitoring over time, to perform periodic updates and optimize the parameters of the digital twin based on new information and evolutions of the physical system under study.
The proposed urban digital twin mainly focuses on specific aspects related to light pollution and hydrogeological risk, but many other applications can be implemented in relation to smart cities. In fact, research is currently focused on identifying functional relationships and models to represent these complex phenomena, which, in themselves, not only require in-depth knowledge of the issues but also optimal and high-performance management of the numerous variables involved.
Despite this, the urban digital twin could certainly include other aspects in the future, such as traffic management, air quality, waste management, and urban safety. As structured, the urban digital twin, with its open-source configuration, could certainly support additional integration across various aspects related to smart cities, helping public administration in different areas of efficient city management. It can also integrate data from various sources (in addition to IoT), including external platforms.
Despite the strengths of the proposed model, some of the main limitations are related to the spatial resolution of VIIRS, which can lead to an underestimation of local illumination in dense urban areas and an overestimation in less illuminated areas. Furthermore, although VIIRS provides data with improved spatial resolution compared to previous satellites, the resolution remains lower than that of other high-resolution sensors. Moreover, VIIRS provides data daily, but daily coverage is not always complete due to atmospheric factors or satellite acquisition times. In some regions, the data may not be available daily due to cloud cover or other atmospheric conditions. Regarding flood risk, despite the quality of data derived from IoT sensors, it can be influenced by the non-uniform distribution of the monitoring stations, which can lead to potential distortions in the results. Therefore, the overall quality of the digital twin may be limited by the availability and quality of geospatial data in different cities.
The proposal aims, first and foremost, to provide concrete and quantifiable answers to these two phenomena, which particularly affect the city of Reggio Calabria, while also leaving room for the implementation of new monitoring and analysis strategies. The digital twin developed is designed to be generalizable and scalable, allowing adaptation to other cities or regions with different geomorphological and infrastructural characteristics. Scalability is ensured using standardized geospatial data from hydraulic and light pollution models based on adaptable parameters. The model’s modular architecture allows for the integration of new data sources. In addition, the calibration methodology can be replicated using local historical data, improving the accuracy of simulations in each application area. The integration of AI and ML techniques, when used independently, represents a powerful resource to enhance predictive analytics and urban resource management. Regarding hydrogeological risk, deep learning can analyze historical rainfall data, hydrometric levels, and satellite imagery to more accurately predict flood events and identify vulnerable areas in real time. Regarding light pollution, clustering algorithms and time series analysis can identify patterns of inefficient energy consumption and suggest strategies for optimizing public lighting. In addition, the use of AI for IoT data analysis enables the processing of large volumes of information in real time, facilitating the creation of dynamic and adaptive scenarios for urban planning. The integration of these advanced tools not only improves the accuracy of simulations but also enables early warning systems and automated solutions for urban resilience.
Moreover, using IoT sensors, satellite imagery, and data from weather stations, the model can be updated in real time to monitor changes in light pollution levels, hydrological conditions, and other critical environmental parameters. This approach allows for early detection of changes in the terrain, prediction of future trends, and support for decisions based on up-to-date data. In the case of hydrogeological risk, continuous monitoring of water levels in basins and rainfall allows early warning systems to be activated, improving emergency management. In the case of light pollution, dynamic monitoring makes it possible to optimize public lighting strategies, reducing energy consumption and environmental impact. In addition, thanks to artificial intelligence and machine learning algorithms, the digital twin can adapt to urban changes, suggesting infrastructure changes and mitigation strategies based on simulated scenarios.

5. Conclusions

The creation of the urban digital twin provides a dynamic and evolving framework for continuous monitoring of various urban processes, such as traffic patterns, pollution levels, and energy consumption. By using the open-source Cesium Ion platform, the system architecture for this urban digital twin is designed to optimize multiple facets of city management, allowing city planners and administrators to make data-driven, informed decisions. The system’s ability to simulate future scenarios further enhances its utility in urban planning, particularly when integrated with emerging technologies like IoT and artificial intelligence. One of the key benefits of such a digital twin system is its capacity to support sustainable urban development. By leveraging real-time data and predictive simulations, municipalities can address challenges related to land management, resource optimization, and environmental impact. Moreover, the ability to simulate different urban scenarios and forecast the potential impacts of various interventions gives administrators valuable insights into the future of the city, helping to guide strategic decision-making processes.
However, while the benefits are considerable, the creation and maintenance of such a complex system come with challenges. The initial investment in tools, technology, and skilled personnel can be significant. Moreover, the sheer volume of data generated—combined with the need for constant updates and real-time processing—presents technical challenges related to data overload, requiring increasingly advanced hardware and software infrastructures. Data security is another pressing concern, especially when handling sensitive information. Therefore, robust security protocols must be implemented to protect the integrity and privacy of the data collected, ensuring that the system is resilient to cyber threats.
Despite these challenges, the research demonstrates the potential of the prototype system, which serves as the foundation for future advancements. Ongoing studies are focused on enhancing the simulation capabilities of the system, integrating artificial intelligence for predictive analysis and further refining the models to provide more accurate insights into the phenomena studied. As the system evolves, it will increasingly incorporate diverse data sources, such as satellite imagery, sensor data, and geospatial analysis, to offer a more comprehensive and detailed view of urban dynamics. Future research will focus on further enhancing the system’s simulation capabilities by integrating advanced artificial intelligence techniques for predictive modeling and real-time adaptation. A key direction will be the incorporation of more diverse and high-resolution data sources, including satellite imagery, IoT sensor networks, and real-time hydrological data, to improve model accuracy and responsiveness. Additionally, efforts will be made to optimize computational efficiency, enabling large-scale simulations applicable to different urban environments. Another important aspect will be the development of adaptive decision-support tools that can assist policymakers in urban planning, flood risk mitigation, and light pollution management. Finally, future studies will explore the potential for scaling the digital twin framework to multiple cities, ensuring generalizability and facilitating cross-regional environmental analysis.
In conclusion, the proposed urban digital twin represents a significant step toward more intelligent, data-driven urban management. While the complexity of the system and the associated challenges require careful consideration, the long-term benefits of improved urban sustainability, risk management, and resource allocation offer a compelling case for further development and refinement of these digital models. The proposed urban digital twin represents a significant advancement toward intelligent, data-driven urban management. By integrating diverse datasets and simulation capabilities, the system enhances decision making for urban sustainability, risk mitigation, and efficient resource allocation. While challenges related to data integration, model calibration, and computational complexity require ongoing attention, the long-term benefits outweigh these obstacles. Future advancements in AI-driven predictive analytics and real-time monitoring will further refine the model, ensuring its adaptability and scalability for diverse urban environments. This research underscores the transformative potential of digital twins in shaping resilient and sustainable cities.

Author Contributions

Conceptualization, V.B., E.G., C.M., S.C., and M.P.M.; methodology, V.B., E.G., C.M., S.C., and M.P.M.; software, V.B., E.G., C.M., S.C., and M.P.M.; validation, V.B., E.G., C.M., S.C., and M.P.M.; formal analysis, V.B., E.G., C.M., S.C., and M.P.M.; investigation, V.B., E.G., C.M., S.C., and M.P.M.; resources, V.B., E.G., C.M., S.C., and M.P.M.; data curation, V.B., E.G., C.M., S.C., and M.P.M.; writing—original draft preparation, V.B., E.G., C.M., S.C., and M.P.M.; writing—review and editing, V.B., E.G., C.M., S.C., and M.P.M.; visualization, V.B., E.G., C.M., S.C., and M.P.M.; supervision, V.B., E.G., C.M., S.C., and M.P.M.; project administration, V.B., E.G., C.M., S.C., and M.P.M.; funding acquisition, V.B., E.G., C.M., S.C., and M.P.M. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the PRIN_2022PNRR project “WebGIS 4D with DSS (Decision Support System) connotation for prediction of landslide susceptibility and hazard through innovative simulation systems with emerging properties such as 3D Cellular Automata, Neural Networks and SPH Fluids” (CUP J53D23019270001) funded by the Italian Ministry of University and Research—MUR (PRIN_2022PNRR_P2022CK8F9).

Data Availability Statement

Contains modified Copernicus Climate Change Service information [2024] available on https://ads.atmosphere.copernicus.eu/datasets/cams-global-emission-inventories?tab=overview accessed on 15 January 2025. Contains modified VIIRS data (doi:10.3390/rs13050922 doi:10.3390/rs13050922) and information accessed with Google Earth Engine. Contains modified Tinitaly DEM data available on https://tinitaly.pi.ingv.it/. Contains modified weather data available on https://open-meteo.com/en/docs/historical-weather-api accessed on 15 January 2025. Contains modified Land Use/Land Cover data available on https://land.copernicus.eu/en/products/urban-atlas/urban-atlas-2018 accessed on 15 January 2025.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Haughton, G. Environmental justice and the sustainable city. In The Earthscan Reader in Sustainable Cities; Routledge: London, UK, 2021; pp. 62–79. [Google Scholar]
  2. Chen, D.; Hu, W.; Li, Y.; Zhang, C.; Lu, X.; Cheng, H. Exploring the temporal and spatial effects of city size on regional economic integration: Evidence from the Yangtze River Economic Belt in China. Land Use Policy 2023, 132, 106770. [Google Scholar] [CrossRef]
  3. Pothukuchi, K. City light or star bright: A review of urban light pollution, impacts, and planning implications. J. Plan. Lit. 2021, 36, 155–169. [Google Scholar] [CrossRef]
  4. Goronczy, E.E. Light Pollution in Metropolises. Analysis, Impacts and Solutions; Springer: Wiesbaden, Germany, 2021. [Google Scholar]
  5. Ding, Y.; Zhang, D.; Lv, J. Comparison of the applicability of city-level building energy consumption quota methods. Energy Build. 2022, 261, 111933. [Google Scholar] [CrossRef]
  6. Rodrigo-Comino, J.; Seeling, S.; Seeger, M.K.; Ries, J.B. Light pollution: A review of the scientific literature. Anthr. Rev. 2023, 10, 367–392. [Google Scholar] [CrossRef]
  7. He, H. Light Pollution Assessment Models Based on Multimodal Machine Learning. Available at SSRN 4511502. 2023. Available online: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4511502 (accessed on 20 February 2025).
  8. Tom, R.O.; George, K.O.; Joanes, A.O.; Haron, A. Review of flood modelling and models in developing cities and informal settlements: A case of Nairobi city. J. Hydrol. Reg. Stud. 2022, 43, 101188. [Google Scholar] [CrossRef]
  9. Globa, A.; Wang, R.; Beza, B. Sensory urbanism and placemaking exploring virtual reality and the creation of place. In Proceedings of the 24th International Conference on Computer-Aided Architectural Design Research in Asia (CAADRIA), Wellington, New Zealand, 15–18 April 2019. [Google Scholar]
  10. Daniel, S.; Doran, M.A. geoSmartCity: Geomatics contribution to the smart city. In Proceedings of the 14th Annual International Conference on Digital Government Research, Quebec City, QC, Canada, 17–20 June 2013; pp. 65–71. [Google Scholar]
  11. González-Zamar, M.D.; Abad-Segura, E.; Vázquez-Cano, E.; López-Meneses, E. IoT technology applications-based smart cities: Research analysis. Electronics 2020, 9, 1246. [Google Scholar] [CrossRef]
  12. Cecchini, C.; Magrini, A.; Gobbi, L. A 3D platform for energy data visualization of building assets. IOP Conf. Ser. Earth Environ. Sci. 2019, 296, 012035. [Google Scholar] [CrossRef]
  13. Chun, P.J.; Yamane, T.; Maharjan, S. Advancing Infrastructure Inspection with AI and 3D Data Platform Integration for Improved Damage Assessment and Enhanced Modeling. In Proceedings of the International Conference on Sustainable Development, Rome, Italy, 11–12 September 2024; Springer Nature: Singapore, 2024; pp. 293–306. [Google Scholar]
  14. Qiu, D.; Lv, B.; Chan, C.M. How digital platforms enhance urban resilience. Sustainability 2022, 14, 1285. [Google Scholar] [CrossRef]
  15. Barrile, V.; Cotroneo, F.; Genovese, E. Platform Prototype for the Prediction of Landslide Susceptibility Through a 4D WebGIS Equipped with Cellular Automata and Neural Networks. In Proceedings of the Italian Conference on Geomatics and Geospatial Technologies, Padua, Italy, 18–20 December 2023; Springer Nature: Cham, Switzerland, 2023; pp. 81–95. [Google Scholar]
  16. Barrile, V.; Genovese, E. Bamboo structures: Innovative methods and applications for structural health monitoring and dissemination. Adv. Bamboo Sci. 2024, 7, 100079. [Google Scholar] [CrossRef]
  17. Barrile, V.; La Foresta, F.; Calcagno, S.; Genovese, E. Innovative System for BIM/GIS Integration in the Context of Urban Sustainability. Appl. Sci. 2024, 14, 8704. [Google Scholar] [CrossRef]
  18. Bilen, T.; Canberk, B.; Duong, T.Q. Digital twin evolution for hard-to-follow aeronautical ad-hoc networks in beyond 5g. IEEE Commun. Stand. Mag. 2023, 7, 4–12. [Google Scholar] [CrossRef]
  19. Bednarz, T.; Baier, A.; Paprocka, I. A framework for communicating and building a digital twin model of the electric car. Appl. Sci. 2024, 14, 1776. [Google Scholar] [CrossRef]
  20. Redeker, M.; Weskamp, J.N.; Rössl, B.; Pethig, F. A digital twin platform for Industrie 4.0. In Data Spaces: Design, Deployment and Future Directions; Springer International Publishing: Cham, Switzerland, 2022; pp. 173–200. [Google Scholar]
  21. Broo, D.G.; Bravo-Haro, M.; Schooling, J. Design and implementation of a smart infrastructure digital twin. Autom. Constr. 2022, 136, 104171. [Google Scholar] [CrossRef]
  22. Pylianidis, C.; Osinga, S.; Athanasiadis, I.N. Introducing digital twins to agriculture. Comput. Electron. Agric. 2021, 184, 105942. [Google Scholar] [CrossRef]
  23. Nasirahmadi, A.; Hensel, O. Toward the next generation of digitalization in agriculture based on digital twin paradigm. Sensors 2022, 22, 498. [Google Scholar] [CrossRef]
  24. Ghandar, A.; Ahmed, A.; Zulfiqar, S.; Hua, Z.; Hanai, M.; Theodoropoulos, G. A Decision Support System for Urban Agriculture Using Digital Twin: A Case Study with Aquaponics. IEEE Access 2021, 9, 35691–35708. [Google Scholar] [CrossRef]
  25. Shengli, W. Is human digital twin possible? Comput. Methods Programs Biomed. Update 2021, 1, 100014. [Google Scholar] [CrossRef]
  26. Song, Y.; Li, Y. Digital twin aided healthcare facility management: A case study of Shanghai tongji hospital. In Proceedings of the Construction Research Congress 2022, Arlington, VA, USA, 9–12 March 2022; pp. 1145–1155. [Google Scholar]
  27. Argota Sánchez-Vaquerizo, J. Getting real: The challenge of building and validating a large-scale digital twin of Barcelona’s traffic with empirical data. ISPRS Int. J. Geo-Inf. 2022, 11, 24. [Google Scholar] [CrossRef]
  28. Lee, A.; Lee, K.W.; Kim, K.H.; Shin, S.W. A geospatial platform to manage large-scale individual mobility for an urban digital twin platform. Remote Sens. 2022, 14, 723. [Google Scholar] [CrossRef]
  29. Xu, H.; Berres, A.; Shao, Y.; Wang, C.R.; New, J.R.; Omitaomu, O.A. Toward a Smart Metaverse City: Immersive Realism and 3D Visualization of Digital Twin Cities. In Advances in Scalable and Intelligent Geospatial Analytics; CRC Press: Boca Raton, FL, USA, 2023; pp. 245–257. [Google Scholar]
  30. Ferré-Bigorra, J.; Casals, M.; Gangolells, M. The adoption of urban digital twins. Cities 2022, 131, 103905. [Google Scholar] [CrossRef]
  31. Mazzetto, S. A Review of Urban Digital Twins Integration, Challenges, and Future Directions in Smart City Development. Sustainability 2024, 16, 8337. [Google Scholar] [CrossRef]
  32. Snap4City. Available online: https://www.snap4city.org/dashboardSmartCity/management/iframeApp.php?linkUrl=https://www.snap4city.org/drupal&linkId=snap4cityPortalLink&pageTitle=www.snap4city.org&fromSubmenu=false (accessed on 15 February 2025).
  33. Schrotter, G.; Hürzeler, C. The digital twin of the city of Zurich for urban planning. PFG–J. Photogramm. Remote Sens. Geoinf. Sci. 2020, 88, 99–112. [Google Scholar] [CrossRef]
  34. Dembski, F.; Wössner, U.; Letzgus, M.; Ruddat, M.; Yamu, C. Urban Digital Twins for Smart Cities and Citizens: The Case Study of Herrenberg, Germany. Sustainability 2020, 12, 2307. [Google Scholar] [CrossRef]
  35. Henriksen, H.J.; Schneider, R.; Koch, J.; Ondracek, M.; Troldborg, L.; Seidenfaden, I.K.; Kragh, S.J.; Bøgh, E.; Stisen, S. A New Digital Twin for Climate Change Adaptation, Water Management, and Disaster Risk Reduction (HIP Digital Twin). Water 2023, 15, 25. [Google Scholar] [CrossRef]
  36. Yu, D.; He, Z. Digital twin-driven intelligence disaster prevention and mitigation for infrastructure: Advances, challenges, and opportunities. Nat. Hazards 2022, 112, 1–36. [Google Scholar] [CrossRef] [PubMed]
  37. Ge, C.; Qin, S. Urban flooding digital twin system framework. Syst. Sci. Control. Eng. 2025, 13, 2460432. [Google Scholar] [CrossRef]
  38. Ariyachandra, M.M.F.; Wedawatta, G. Digital twin smart cities for disaster risk management: A review of evolving concepts. Sustainability 2023, 15, 11910. [Google Scholar] [CrossRef]
  39. Shahat, E.; Hyun, C.T.; Yeom, C. City Digital Twin Potentials: A Review and Research Agenda. Sustainability 2021, 13, 3386. [Google Scholar] [CrossRef]
  40. Koeva, M.; Tripathy, A.; Afzalinezhad, A. The Role of Digital Twins in Mitigating Urban Heat Islands: How Cities Are Leveraging Innovative Technology for Sustainable Urban Development; GIM International: Lemmer, The Netherlands, 2024. [Google Scholar]
  41. Chen, C.; Han, Y.; Galinski, A.; Calle, C.; Carney, J.; Ye, X.; van Westen, C. Integrating urban digital twins with cloud-based geospatial dashboards for coastal resilience planning: A case study in Florida. arXiv 2024, arXiv:2403.18188. [Google Scholar] [CrossRef]
  42. Elvidge, C.D.; Zhizhin, M.; Ghosh, T.; Hsu, F.C.; Taneja, J. Annual time series of global VIIRS nighttime lights derived from monthly averages: 2012 to 2019. Remote Sens. 2021, 13, 922. [Google Scholar] [CrossRef]
  43. Agustí-Panareda, A.; Massart, S.; Chevallier, F.; Balsamo, G.; Boussetta, S.; Dutra, E.; Beljaars, A. A biogenic CO2 flux adjustment scheme for the mitigation of large-scale biases in global atmospheric CO2 analyses and forecasts. Atmos. Chem. Phys. 2016, 16, 10399–10418. [Google Scholar] [CrossRef]
  44. DJI Enterprise. Available online: https://enterprise.dji.com/it/matrice-350-rtk/specs (accessed on 3 January 2024).
  45. Tarquini, S.; Isola, M.I.; Favalli, A.; Battistini, G.D. TINITALY, a Digital Elevation Model of Italy with a 10 Meters Cell Size (Version 1.1); Istituto Nazionale di Geofisica e Vulcanologia (INGV): Rome, Italy, 2023. [CrossRef]
  46. Tarquini, S.; Isola, I.; Favalli, M.; Mazzarini, F.; Bisson, M.; Pareschi, M.T.; Boschi, E. TINITALY/01: A new triangular irregular network of Italy. Ann. Geophys. 2007, 50, 407–425. [Google Scholar] [CrossRef]
  47. Available online: https://www.sandinginstrument.com/en/product_con.html?top=GNSS&c=RTKReceivers&short_url=AquaT5#:~:text=Aqua%20T5%20is%20a%20new%20product%20from%20SANDING,the%20ability%20of%20receiving%20higher%20quality%20satellites%20signals (accessed on 14 February 2024).
  48. Zippenfenig, P. Open-Meteo.com Weather API [Computer Software]; Zenodo: Geneva, Switzerland, 2023. [Google Scholar] [CrossRef]
  49. Urban Atlas Land Cover/Land Use 2018 (Vector), Europe. [Data Set]. November 2024. Available online: https://land.copernicus.eu/en/products/urban-atlas/urban-atlas-2018 (accessed on 17 February 2024).
  50. Cesium. Available online: https://cesium.com/platform/cesium-ion/ (accessed on 3 January 2024).
  51. Pellegrino, A.; Borrelli, S. 23. Analisi del Dissesto da Frana in Calabria; APAT (Agenzia per la Protezione dell’Ambiente e per i Servizi Tecnici): Rome, Italy, 2007. [Google Scholar]
  52. Ferro-Escobar, R.; Vacca-González, H.; Gómez-Castillo, H. Smart and Sustainable Cities in Collaboration with IoT: The Singapore Success Case. In Machine Learning for Smart Environments/Cities; Intelligent Systems Reference Library; Marques, G., González-Briones, A., Molina López, J.M., Eds.; Springer: Cham, Switzerland, 2022; Volume 121. [Google Scholar] [CrossRef]
Futureinternet 17 00110 g001
Figure 1. Research flow.
Figure 1. Research flow.
Futureinternet 17 00110 g001
Futureinternet 17 00110 g002
Figure 2. Sensors’ acquisition system: (a) complete system for the acquisition of sensor data; (b) part of the sensors installed in the city center of Reggio Calabria.
Figure 2. Sensors’ acquisition system: (a) complete system for the acquisition of sensor data; (b) part of the sensors installed in the city center of Reggio Calabria.
Futureinternet 17 00110 g002
Futureinternet 17 00110 g003
Figure 3. Urban digital twin architecture.
Figure 3. Urban digital twin architecture.
Futureinternet 17 00110 g003
Futureinternet 17 00110 g004
Figure 4. Study area; Municipality of Reggio Calabria (Italy).
Figure 4. Study area; Municipality of Reggio Calabria (Italy).
Futureinternet 17 00110 g004
Futureinternet 17 00110 g005
Figure 5. Preview of the urban digital twin of Reggio Calabria: 3D representation of light pollution expressed in radiative flux.
Figure 5. Preview of the urban digital twin of Reggio Calabria: 3D representation of light pollution expressed in radiative flux.
Futureinternet 17 00110 g005
Futureinternet 17 00110 g006
Figure 6. Preview of the urban digital twin of Reggio Calabria: 3D representation of hydrogeological risk.
Figure 6. Preview of the urban digital twin of Reggio Calabria: 3D representation of hydrogeological risk.
Futureinternet 17 00110 g006
Futureinternet 17 00110 g007
Figure 7. Correlation matrix between light pollution and CO2.
Figure 7. Correlation matrix between light pollution and CO2.
Futureinternet 17 00110 g007
Futureinternet 17 00110 g008
Figure 8. Correlation matrix between land use, risk area, rainfall, and slope.
Figure 8. Correlation matrix between land use, risk area, rainfall, and slope.
Futureinternet 17 00110 g008
Futureinternet 17 00110 g009
Figure 9. Distribution of hazard areas by land use class.
Figure 9. Distribution of hazard areas by land use class.
Futureinternet 17 00110 g009
Table 1. Data sources.
Table 1. Data sources.
DataSource
VIIRS Nighttime Lights (VNL)https://ladsweb.modaps.eosdis.nasa.gov/missions-and-measurements/viirs/, (accessed on 15 February 2025) [42]
CAMS Global Emission Inventorieshttps://ads.atmosphere.copernicus.eu/datasets/cams-global-emission-inventories?tab=overview, (accessed on 15 February 2025) [43]
Tinitaly DEMhttps://tinitaly.pi.ingv.it/, (accessed on 15 February 2025) [45,46]
Light sources positionsRover GNSS Sanding T5 [47]
Precipitation datahttps://open-meteo.com/en/docs/historical-weather-api, (accessed on 15 February 2025) [48]
Urban Atlashttps://land.copernicus.eu/en/products/urban-atlas/urban-atlas-2018, (accessed on 15 February 2025) [49]
Points cloud from drone equipped with LiDAR sensorSurvey with DJI Matrice 350 RTK drone [44] equipped with a LiDAR sensor
Table 2. High light pollution area percentage in land use class.
Table 2. High light pollution area percentage in land use class.
Land Use ClassHigh Light Pollution Area Percentage
11100 Continuous Urban Fabric33.3%
11210 Discontinuous Dense Urban Fabric16.67%
11220 Discontinuous Medium Density Urban Fabric2.13%
11230 Discontinuous Low Density Urban Fabric2.08%
11300 Isolated Structures1.92%
12100 Industrial, Commercial, Public, Military, and Private Units0.88%
12210 Fast Transit Roads and Associated Land13.33%
12220 Other Roads and Associated Land1.45%
12230 Railways and Associated Land5%
12300 Port Areas11%
13100 Mineral Extraction and Dump Sites1.11%
13400 Low without Current Use0.72%
14100 Green Urban Areas1.27%
14200 Sports and Leisure Facilities0.4%
22000 Permanent Crops1.69%
23000 Pastures0.75%
31000 Forests4.35%
32000 Herbaceous Vegetation Associations1.39%
33000 Open Spaces with Little or No Vegetation0.53%
Table 3. Risk area percentage in land use class.
Table 3. Risk area percentage in land use class.
Land Use ClassRisk Area Percentage
11100 Continuous Urban Fabric3.85%
11210 Discontinuous Dense Urban Fabric4.78%
11220 Discontinuous Medium Density Urban Fabric3.61%
11230 Discontinuous Low Density Urban Fabric0.72%
11300 Isolated Structures0.21%
12100 Industrial, Commercial, Public, Military, and Private Units3.49%
12210 Fast Transit Roads and Associated Land0.40%
12220 Other Roads and Associated Land9.12%
12230 Railways and Associated Land0.20%
12300 Port Areas3.42%
13100 Mineral Extraction and Dump Sites1.87%
13400 Low without Current Use0.59%
14100 Green Urban Areas0.68%
14200 Sports and Leisure Facilities1.34%
22000 Permanent Crops7.94%
23000 Pastures39.58%
31000 Forests0.50%
32000 Herbaceous Vegetation Associations15.76%
33000 Open Spaces with Little or No Vegetation1.94%
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Barrile, V.; Genovese, E.; Maesano, C.; Calluso, S.; Manti, M.P. Developing an Urban Digital Twin for Environmental and Risk Assessment: A Case Study on Public Lighting and Hydrogeological Risk. Future Internet 2025, 17, 110. https://doi.org/10.3390/fi17030110

AMA Style

Barrile V, Genovese E, Maesano C, Calluso S, Manti MP. Developing an Urban Digital Twin for Environmental and Risk Assessment: A Case Study on Public Lighting and Hydrogeological Risk. Future Internet. 2025; 17(3):110. https://doi.org/10.3390/fi17030110

Chicago/Turabian Style

Barrile, Vincenzo, Emanuela Genovese, Clemente Maesano, Sonia Calluso, and Maurizio Pasquale Manti. 2025. "Developing an Urban Digital Twin for Environmental and Risk Assessment: A Case Study on Public Lighting and Hydrogeological Risk" Future Internet 17, no. 3: 110. https://doi.org/10.3390/fi17030110

APA Style

Barrile, V., Genovese, E., Maesano, C., Calluso, S., & Manti, M. P. (2025). Developing an Urban Digital Twin for Environmental and Risk Assessment: A Case Study on Public Lighting and Hydrogeological Risk. Future Internet, 17(3), 110. https://doi.org/10.3390/fi17030110

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop