Insights into susceptibility of underground infrastructure to geohazards due to subsurface urban heat Island

Bibliographic Data

ID21228296
AuthorsMary C Ngoma (0000-0001-9211-482X, New Jersey Institute of Technology), Oladoyin Kolawole (0000-0001-9202-172X, New Jersey Institute of Technology, corresponding author), Mohsen Zaker Esteghamati (0000-0002-2144-2938, Utah State University)
Year2025
Volume125
Pages106332
Publication date2025-05-01
Open AccessYes
TypeARTICLE
VenueSustainable Cities and Society (JOURNAL)
Journal identifiersISSN: 2210-6707 • E-ISSN: 2210-6715
PublisherElsevier BV (PUBLISHER)
DOI10.1016/j.scs.2025.106332
OpenAlexW4408863535
LanguageEN
References cited47

Proposed numerical framework to assess the spatio-temporal evolution of SUHI. • SVM emerged as the best algorithm for predicting SUHI underground impact. • SUHI's impact on underground conditions has intensified over the past decade. • Drastic drop in model accuracy with most recent field data, due to concept drift effect. • First reveal of a link between concept drift and SUHI. Subsurface urban heat island (SUHI) is a global phenomenon caused by warming up of the underground at a fast rate in densely-populated urban areas due to heat dissipation from surface structures. Exploring the potential hazards in the underground induced by SUHI has become crucial for developing resilient urban infrastructure with excess-heat mitigation measures. However, despite increasing studies on urban heat island, there is currently a knowledge gap in SUHI on the performance of underground soil, which may temporally result in geohazards and failure of surface and subsurface infrastructure. This study proposes a numerical framework that incorporates critical field datasets to investigate the spatio-temporal evolution of SUHI impact on the underground in urban environments, and the extent to which SUHI can trigger future detrimental impact. We further assessed the resultant potential susceptibility of subsurface infrastructure in built-environment if trends continue in fast-growing urban cities. The results show that the best-performing algorithm developed for predicting the underground impact of SUHI is Support Vector Machine (SVM) with the lowest error values (MAE = 6.43 tsf, MAPE = 11 %, MSE = 183 tsf2) and the highest R 2 score of 0.77. However, when the model is subjected to the most recent field data from the analyzed city, a drastic reduction in accuracy is observed, suspected to be caused by a phenomenon known as “ concept drift ”. As underground warming increases due to temperature variations induced by SUHI, these may represent a potential hazard for urban cities globally. This novel finding of historical non-conformance yielding the concept drift highlights the complex and dynamic nature of SUHI which necessitates the development of more resilient urban infrastructure and sustainable cities. This study also provides valuable insights into the need for more resilient urban infrastructures that can mitigate SUHI risks and ensure subsurface infrastructure remains sustainable and adaptable to emerging challenges

Civil engineering · Earth science · Environmental planning · Geography · Meteorology · Mining engineering · Urban heat island · Climate change and permafrost · Engineering · Environmental Science · Landslides and related hazards · Seismic Waves and Analysis · Geology