Downscaling mapping method for local climate zones from the perspective of deep learning
文献类型:期刊论文
作者 | Yu, Wenbo; Yang, Jun5,6; Wu, Feng1; He, Baojie7; Yu, Huisheng; Ren, Jiayi; Xiao, Xiangming4; Xia, Jianhong(Cecilia) |
刊名 | URBAN CLIMATE |
出版日期 | 2023-05-01 |
卷号 | 49页码:101500 |
ISSN号 | 2212-0955 |
关键词 | Local climate zone Refinement mapping Optimal mapping scale analysis Deep learning model optimization Geographic information system |
DOI | 10.1016/j.uclim.2023.101500 |
文献子类 | Article |
英文摘要 | Increased attention has been paid to refining Local Climate Zone (LCZ) maps. However, the mapping process is difficult because of image resolution limitations and the lack of detailed planning data. To our knowledge for the first time, to promote the application of refined LCZ maps in urban climate research, we have proposed the downscaling mapping process by by integrating GIS workflow and remote sensing workflow. The results have shown that: (1) By adjusting the scale and input of the deep learning model, the LCZ basemap achieved the highest overall accuracy of 83.64% and the highest data matching degree of 0.71. (2) Visual interpre-tation showed that the overall accuracy of LCZ map at all scales is >70%, and the highest overall accuracy of 84.44% is achieved at 20 m scale. (3) Usefulness assessment showed that the LST levels of LCZs significantly differ at all scales (p < 0.01). The 160 m scale was the most suitable for analysis of the thermal characteristics of urban landscapes. This downscaling mapping method will help urban planners in developing urban climate models to determine the impact mecha-nisms of urban development and support progress in sustainable urban development efforts. |
WOS关键词 | LAND-SURFACE TEMPERATURE ; URBAN HEAT-ISLAND ; IMAGE-ANALYSIS ; CLASSIFICATION ; NETWORK ; WUDAPT ; SCHEME ; AREAS |
WOS研究方向 | Environmental Sciences & Ecology ; Meteorology & Atmospheric Sciences |
出版者 | ELSEVIER |
WOS记录号 | WOS:000960511200001 |
源URL | [http://ir.igsnrr.ac.cn/handle/311030/190477] |
专题 | 区域可持续发展分析与模拟院重点实验室_外文论文 |
作者单位 | 1.Liaoning Normal Univ, Human Settlements Res Ctr, Dalian 116029, Peoples R China 2.Xia, Jianhong(Cecilia)] Curtin Univ, Sch Earth & Planetary Sci EPS, Perth 65630, Australia 3.Univ Oklahoma, Ctr Earth Observat & Modeling, Dept Microbiol & Plant Biol, Norman, OK 73019 USA 4.Chongqing Univ, Ctr Climate Resilient & Low Carbon Cities, Sch Architecture & Urban Planning, Chongqing 400045, Peoples R China 5.Northeastern Univ, Sch Humanities & Law, Shenyang 110169, Peoples R China 6.Northeastern Univ, Jangho Architecture Coll, Shenyang 110169, Peoples R China 7.Chinese Acad Sci, Inst Geog Sci & Nat Resources Res, Key Lab Reg Sustainable Dev Modeling, Beijing 100101, Peoples R China |
推荐引用方式 GB/T 7714 | Yu, Wenbo,Yang, Jun,Wu, Feng,et al. Downscaling mapping method for local climate zones from the perspective of deep learning[J]. URBAN CLIMATE,2023,49:101500. |
APA | Yu, Wenbo.,Yang, Jun.,Wu, Feng.,He, Baojie.,Yu, Huisheng.,...&Xia, Jianhong.(2023).Downscaling mapping method for local climate zones from the perspective of deep learning.URBAN CLIMATE,49,101500. |
MLA | Yu, Wenbo,et al."Downscaling mapping method for local climate zones from the perspective of deep learning".URBAN CLIMATE 49(2023):101500. |
入库方式: OAI收割
来源:地理科学与资源研究所
浏览0
下载0
收藏0
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。