Air pollution is a global problem, which needs to be understood and controlled to ensure a healthy environment and inform sustainable development. Urban areas have been established as one of the main contributors to air pollution, and, as such, urban air quality is the subject of an increasing volume of research. One of the principal means of studying air pollution dispersion is to use computational fluid dynamics (CFD) models. Subject to careful verification and validation, these models allow for analysts to predict air flow and pollution concentration for various urban morphologies under different environmental conditions. This article presents a detailed review of the use of CFD to model air pollution dispersion in an urban environment over the last decade. The review extracts and summarises information from nearly 90 pieces of published research, categorising it according to over 190 modelling features, which are thematically systemised into 7 groups. The findings from across the field are critically compared to available urban air pollution modelling guidelines and standards. Among the various quantitative trends and statistics from the review, two key findings stand out. The first is that, despite the existence of best practice guidelines for pollution dispersion modelling, anywhere between 12% and 34% of the papers do not specify one or more aspects of the utilised models, which are required to reproduce the study. The second is that none of the articles perform verification and validation according to accepted standards. The results of this review can, therefore, be used by practitioners in the field of pollution dispersion modelling to understand the general trends in current research and to identify open problems to be addressed in the future.
People tend to spend the majority of their time indoors. Indoor air properties can significantly affect humans’ comfort, health, and productivity. This study utilizes measurement data of indoor conditions in a kindergarten in Sofia, Bulgaria. Autoregressive integrated moving average (ARIMA) and long short-term memory (LSTM) recurrent neural network (RNN) models were developed to predict CO levels in the educational facility over the next hour based on 2.5 h of past data and allow for near real-time decision-making. The better-performing model, LSTM, is also used for temperature and relative humidity forecasting. Global comfort is then estimated based on threshold values for temperature, humidity, and CO. The predicted values ranged between 0.938 and 0.981 for the three parameters, while the prediction of global comfort conditions achieved a 91/100 accuracy.
Accurate prediction of urban wind flow is essential for urban planning and environmental assessment. Classical computational fluid dynamics (CFD) methods are computationally expensive, while machine learning approaches often lack explainability and generalizability. To address the limitations of both approaches, we propose Diff-KNN, a hybrid method that combines Coarse-Scale CFD simulations with a K-Nearest Neighbors (KNN) model trained on the residuals between coarse- and fine-scale CFD results. Diff-KNN reduces velocity prediction errors by up to 83.5% compared to Pure-KNN and 56.6% compared to coarse CFD alone. Tested on the AIJE urban dataset, Diff-KNN effectively corrects flow inaccuracies near buildings and within narrow street canyons, where traditional methods struggle. This study demonstrates how residual learning can bridge physics-based and data-driven modeling for accurate and interpretable fine-scale urban wind prediction.
Computational fluid dynamics (CFD) models, used independently or integrated into city digital twins, can predict air pollution dispersion and support informed decision-making in urban planning. While modelling complexity and capabilities have increased, systematic validation and uncertainty quantification remain crucial, yet are among the least understood and applied aspects of the analysis process. To address this issue, the ValUr project, supported by the ERIES programme, aims to establish a best-practice protocol for urban CFD validation and to generate a high-quality, open-access pollution dispersion dataset. This work introduces the project scope, motivation, and goals, detailing early project activities, including the identification of the study area - part of the city centre of Sofia, Bulgaria, where air quality is a pressing concern, and the creation of a scaled geometry suitable for physical model-making, wind tunnel testing, and CFD simulations. The geometry goes beyond the basic level of detail (LoD) 1, typically used in validation databases and reaches LoD 2.2, capturing the complex building morphologies and realistic urban conditions of the area. By sharing protocols and data openly, ValUr seeks to promote the importance of model reliability and to encourage validation consistency within the computational wind engineering community.
Cultural heritage buildings constitute an irreplaceable record of historical, social, and architectural identity, and their preservation is essential for cultural continuity and sustainable development. However, their conservation is inherently challenging due to material aging, complex construction techniques, limited documentation, and strict intervention constraints that restrict invasive monitoring or retrofitting solutions. Environmental degradation and microclimatic effects further accelerate deterioration, often in ways that are difficult to quantify or predict. This paper presents RADAR, a non-invasive, data-driven framework for assessing environmental and structural risk in cultural heritage buildings. The proposed approach integrates high-resolution geometric acquisition, computational fluid dynamics (CFD) modeling, and environmental monitoring to analyze airflow patterns, temperature distribution, and moisture-related decay mechanisms. By combining measured data with numerical simulations, RADAR enables the identification of high-risk zones and deterioration drivers without altering the building fabric. The framework is demonstrated through a real-world case study, illustrating its applicability as a decision-support tool for preventive conservation and heritage management.