{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,31]],"date-time":"2025-10-31T07:12:14Z","timestamp":1761894734203,"version":"build-2065373602"},"reference-count":18,"publisher":"MDPI AG","issue":"4","license":[{"start":{"date-parts":[[2025,10,29]],"date-time":"2025-10-29T00:00:00Z","timestamp":1761696000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["MAKE"],"abstract":"<jats:p>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.<\/jats:p>","DOI":"10.3390\/make7040131","type":"journal-article","created":{"date-parts":[[2025,10,29]],"date-time":"2025-10-29T17:38:22Z","timestamp":1761759502000},"page":"131","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Diff-KNN: Residual Correction of Baseline Wind Predictions in Urban Settings"],"prefix":"10.3390","volume":"7","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-3837-9461","authenticated-orcid":false,"given":"Dimitri","family":"Nowak","sequence":"first","affiliation":[{"name":"Fraunhofer Institute for Industrial Mathematics ITWM, Optimization Department, Fraunhofer-Platz 1, 67663 Kaiserslautern, Germany"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4513-5035","authenticated-orcid":false,"given":"Jennifer","family":"Werner","sequence":"additional","affiliation":[{"name":"Fraunhofer Institute for Industrial Mathematics ITWM, Optimization Department, Fraunhofer-Platz 1, 67663 Kaiserslautern, Germany"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2375-1328","authenticated-orcid":false,"given":"Franziska","family":"Hunger","sequence":"additional","affiliation":[{"name":"Fraunhofer-Chalmers Centre for Industrial Mathematics, Computational Engineering and Design Department, Chalmers Science Park, SE-412 88 Gothenburg, Sweden"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4472-5700","authenticated-orcid":false,"given":"Tomas","family":"Johnson","sequence":"additional","affiliation":[{"name":"Fraunhofer-Chalmers Centre for Industrial Mathematics, Computational Engineering and Design Department, Chalmers Science Park, SE-412 88 Gothenburg, Sweden"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0038-3307","authenticated-orcid":false,"given":"Andreas","family":"Mark","sequence":"additional","affiliation":[{"name":"Fraunhofer-Chalmers Centre for Industrial Mathematics, Computational Engineering and Design Department, Chalmers Science Park, SE-412 88 Gothenburg, Sweden"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2129-9860","authenticated-orcid":false,"given":"Radostin","family":"Mitkov","sequence":"additional","affiliation":[{"name":"\u201cBig Data for Smart Society\u201d (GATE) Institute, Sofia University \u201cSt. Kliment Ohridski\u201d, 5 James Bourchier Blvd., 1164 Sofia, Bulgaria"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3894-1837","authenticated-orcid":false,"given":"Fredrik","family":"Edelvik","sequence":"additional","affiliation":[{"name":"Fraunhofer-Chalmers Centre for Industrial Mathematics, Computational Engineering and Design Department, Chalmers Science Park, SE-412 88 Gothenburg, Sweden"}]}],"member":"1968","published-online":{"date-parts":[[2025,10,29]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"107397","DOI":"10.1016\/j.buildenv.2020.107397","article-title":"Non-intrusive reduced order model of urban airflow with dynamic boundary conditions","volume":"187","author":"Xiang","year":"2021","journal-title":"Build. Environ."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"100941","DOI":"10.1016\/j.uclim.2021.100941","article-title":"Fast simulation of high resolution urban wind fields at city scale","volume":"39","author":"Xiang","year":"2021","journal-title":"Urban Clim."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"108315","DOI":"10.1016\/j.buildenv.2021.108315","article-title":"Deep learning to replace, improve, or aid CFD analysis in built environment applications: A review","volume":"206","author":"Calzolari","year":"2021","journal-title":"Build. Environ."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"154223","DOI":"10.1016\/j.scitotenv.2022.154223","article-title":"Combining computational fluid dynamics and neural networks to characterize microclimate extremes: Learning the complex interactions between meso-climate and urban morphology","volume":"829","author":"Javanroodi","year":"2022","journal-title":"Sci. Total Environ."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"109575","DOI":"10.1016\/j.buildenv.2022.109575","article-title":"Accelerated environmental performance-driven urban design with generative adversarial network","volume":"224","author":"Huang","year":"2022","journal-title":"Build. Environ."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"113622","DOI":"10.1016\/j.buildenv.2025.113622","article-title":"Accelerating Large Eddy Simulations of Urban Airflow with Generative Adversarial Networks","volume":"286","author":"Calzolari","year":"2025","journal-title":"Build. Environ."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"111063","DOI":"10.1016\/j.buildenv.2023.111063","article-title":"Fourier neural operator for real-time simulation of 3D dynamic urban microclimate","volume":"248","author":"Peng","year":"2024","journal-title":"Build. Environ."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"110056","DOI":"10.1016\/j.buildenv.2023.110056","article-title":"PIGNN-CFD: A physics-informed graph neural network for rapid predicting urban wind field defined on unstructured mesh","volume":"232","author":"Shao","year":"2023","journal-title":"Build. Environ."},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"BenMoshe, N., Fattal, E., Leitl, B., and Arav, Y. (2023). Using Machine Learning to Predict Wind Flow in Urban Areas. Atmosphere, 14.","DOI":"10.3390\/atmos14060990"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"105387","DOI":"10.1016\/j.envsoft.2022.105387","article-title":"Hazardous atmospheric dispersion in urban areas: A Deep Learning approach for emergency pollution forecast","volume":"152","author":"Mendil","year":"2022","journal-title":"Environ. Model. Softw."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"112229","DOI":"10.1016\/j.buildenv.2024.112229","article-title":"Machine Learning to speed up Computational Fluid Dynamics engineering simulations for built environments: A review","volume":"267","author":"Caron","year":"2025","journal-title":"Build. Environ."},{"key":"ref_12","unstructured":"AIJ (2016). AIJ Benchmarks for Validation of CFD Simulations Applied to Pedestrian Wind Environment Around Buildings, Architectural Institute of Japan (AIJ)."},{"key":"ref_13","first-page":"241","article-title":"Comparison Between Different Immersed Boundary Conditions for Simulation of Complex Fluid Flows","volume":"7","author":"Mark","year":"2011","journal-title":"Fluid Dyn. Mater. Process."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"112036","DOI":"10.1016\/j.buildenv.2024.112036","article-title":"Evaluation of an immersed boundary numerical framework to address the wind field in complex urban topographies","volume":"266","author":"Vanky","year":"2024","journal-title":"Build. Environ."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"108812","DOI":"10.1016\/j.engappai.2024.108812","article-title":"Optimisation of city structures with respect to high wind speeds using U-Net models","volume":"135","author":"Nowak","year":"2024","journal-title":"Eng. Appl. Artif. Intell."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"98","DOI":"10.3390\/make6010006","article-title":"Predicting Wind Comfort in an Urban Area: A Comparison of a Regression- with a Classification-CNN for General Wind Rose Statistics","volume":"6","author":"Werner","year":"2024","journal-title":"Mach. Learn. Knowl. Extr."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"012060","DOI":"10.1088\/1742-6596\/3027\/1\/012060","article-title":"On the variety and adequacy of different solution verification approaches in computational wind engineering","volume":"3027","author":"Mitkov","year":"2025","journal-title":"J. Phys. Conf. Ser."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"265","DOI":"10.1016\/j.jcp.2004.10.018","article-title":"Near-wall behavior of RANS turbulence models and implications for wall functions","volume":"204","author":"Kalitzin","year":"2005","journal-title":"J. Comput. Phys."}],"container-title":["Machine Learning and Knowledge Extraction"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2504-4990\/7\/4\/131\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,31]],"date-time":"2025-10-31T05:34:40Z","timestamp":1761888880000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2504-4990\/7\/4\/131"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,29]]},"references-count":18,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2025,12]]}},"alternative-id":["make7040131"],"URL":"https:\/\/doi.org\/10.3390\/make7040131","relation":{},"ISSN":["2504-4990"],"issn-type":[{"type":"electronic","value":"2504-4990"}],"subject":[],"published":{"date-parts":[[2025,10,29]]}}}