Prediction of urban heat island intensity based on multiple linear regression and deep learning
PLOS One · 29.12.2025 · 20 (12) · e0339782
DOI: 10.1371/journal.pone.0339782 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS One, online erschienen am 29.12.2025. Verfasst von Xinran Liu und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Remote Sensing Information Engineering, North China Institute of Aerospace Engineering, Lang fan, He Bei, China. Hauptthemen laut MEDLINE: Deep Learning, Hot Temperature. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The rapid urbanization process has led to many prominent environmental issues in urban areas, resulting from a drastic change in land use. The Urban Heat Island (UHI) effect is of particular concern because it has a significant impact on the livability of cities. Therefore, exploring and studying the intensity of UHI and its future distribution have significant practical implications. In this study, the CNN-LSTM-Attention model was constructed to predict four remote sensing spectral indices, and combined with the CA-Markov model to predict land use change. The relationship between four different spectral indices and the intensity of UHI was analyzed by a multiple linear regression model (R 2 = 0.7468, RMSE = 0.0546), and the UHI intensity and distribution in 2025 were predicted and analyzed. The results show that by 2025, the proportion of built-up area will continue to increase by 1.37 percentage points, which will lead to a more intense and concentrated UHI effect, and the proportion of heat island area will increase by 2.97 percentage points. The study shows that increasing vegetation area and water area can effectively alleviate the impact of UHI. Local government departments formulate reasonable policies based on survey results, reduce over-radiation, improve urban livability, and promote sustainable development.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Remote Sensing Information Engineering, North China Institute of Aerospace Engineering, Lang fan, He Bei, China.
ORCID-Profile: Xia Zhu
Zitieren
APA
Liu, X., Zhu, X., Liu, Y., Jia, C., Cao, J., Zhong, Y., Zhang, Y., & Zhang, Y. (2025). Prediction of urban heat island intensity based on multiple linear regression and deep learning. PLOS One, 20(12), e0339782. https://doi.org/10.1371/journal.pone.0339782
Vancouver
Liu X, Zhu X, Liu Y, Jia C, Cao J, Zhong Y, et al. Prediction of urban heat island intensity based on multiple linear regression and deep learning. PLOS One. 2025;20(12):e0339782. doi:10.1371/journal.pone.0339782
BibTeX
@article{Liu2025,
author = {Liu, Xinran and Zhu, Xia and Liu, Yuanping and Jia, Cui and Cao, Jie and Zhong, Yu and Zhang, Yu and Zhang, Yunjie},
title = {Prediction of urban heat island intensity based on multiple linear regression and deep learning},
journal = {PLOS One},
year = {2025},
volume = {20},
number = {12},
pages = {e0339782},
doi = {10.1371/journal.pone.0339782}
}
EUVIMED