气候变化领域集成服务门户(CCPortal): Spatial-temporal analysis of atmospheric environment in urban areas using remote sensing and neural networks

CCPortal

DOI	10.1016/j.suscom.2024.100987
	Spatial-temporal analysis of atmospheric environment in urban areas using remote sensing and neural networks
	Mokarram, Marzieh; Taripanah, Farideh; Pham, Tam Minh
发表日期	2024
ISSN	2210-5379
EISSN	2210-5387
起始页码	42
卷号	42
英文摘要	Rapid urbanization has given rise to escalating land surface temperatures, climate change, and the emergence of surface urban heat islands (SUHIs) and urban hot spots (UHSs), posing significant environmental challenges. This study, situated in the dynamic urban landscape of southern Iran, leverages Landsat satellite imagery to scrutinize the repercussions of temperature escalation on the environment. Our approach harnesses a novel Urban Thermal Field Variance Index (UTFVI) in conjunction with thermal and spectral indices to gain insights into these challenges. We employ a multifaceted methodology that integrates linear regression, cellular automata (CA)Markov chains, and advanced neural network techniques to predict land surface temperature (LST) values and associated indicators. Over the span of 2000 - 2019, our findings reveal a 5% augmentation in urban heat islands (UHIs), signifying an alarming temperature increase. A striking 46% of the region, as uncovered by UTFVI, falls into the most severe categories of ecological discomfort. Our analysis underscores the robust correlations between LST and critical indices, notably the Normalized Difference Built Index (NDBI) (0.96), Normalized Difference Vegetation Index (NDVI) (-0.71), UTFVI (0.98), and SUHI (0.82). Notably, our original contributions lie in the application of Artificial Neural Networks (ANNs), wherein the Multilayer Perceptron (MLP) method excels in predicting UTFVI (R 2 =0.96) and NDBI (R 2 =0.96), while the Radial Basis Function (RBF) method demonstrates remarkable accuracy in forecasting the SUHI index (R 2 =0.96). These achievements signify a groundbreaking advancement in comprehending the intricate dynamics of urban environmental conditions. The repercussions of increased urbanization, the proliferation of barren land, and dwindling vegetation in 2019 manifest in a marked decline in ecological quality, with a concomitant surge in temperatures within the study area. These findings underscore the pressing need for informed urban planning and sustainable practices to mitigate the detrimental effects of urban heat islands and their impact on local climates.
英文关键词	Urbanization; Atmospheric Environment; Remote Sensing; Neural Networks
语种	英语
WOS研究方向	Computer Science
WOS类目	Computer Science, Hardware & Architecture ; Computer Science, Information Systems
WOS记录号	WOS:001235069900001
来源期刊	SUSTAINABLE COMPUTING-INFORMATICS & SYSTEMS
文献类型	期刊论文
条目标识符	http://gcip.llas.ac.cn/handle/2XKMVOVA/302411
作者单位	Shiraz University; University Kashan; Vietnam National University Hanoi; Vietnam National University Hanoi
推荐引用方式 GB/T 7714	Mokarram, Marzieh,Taripanah, Farideh,Pham, Tam Minh. Spatial-temporal analysis of atmospheric environment in urban areas using remote sensing and neural networks[J],2024,42.
APA	Mokarram, Marzieh,Taripanah, Farideh,&Pham, Tam Minh.(2024).Spatial-temporal analysis of atmospheric environment in urban areas using remote sensing and neural networks.SUSTAINABLE COMPUTING-INFORMATICS & SYSTEMS,42.
MLA	Mokarram, Marzieh,et al."Spatial-temporal analysis of atmospheric environment in urban areas using remote sensing and neural networks".SUSTAINABLE COMPUTING-INFORMATICS & SYSTEMS 42(2024).