ScholarMate
English简体中文繁體中文
客服热线:400-1616-289
登录注册
期刊论文

Remote sensing estimation of surface PM2.5 concentrations using a deep learning model improved by data augmentation and a particle size constraint

Yin, Shunchao; Li, Tongwen*; Cheng, Xiao*; Wu, Jingan中山大学
Atmospheric Environment2022-10-15DOI: 10.1016/j.atmosenv.2022.119282
17被引13阅读
引用收藏
分享

摘要

Accurate estimation of PM2.5 concentrations is critical to understanding and counteracting air pollution. In the past decade, various machine learning models, especially deep learning models, have been widely used in PM2.5 remote sensing estimation and have achieved remarkable performance. However, a typical pitfall of deep learning models is the problem of high-value underestimation, i.e., the models often underestimate high-level PM2.5 concentrations. Alleviating the problem of high-value underestimation and improving the estimation ac-curacy of high PM2.5 concentrations are crucial. This study developed a new deep learning model that combines data augmentation and a particle size constraint to improve the high-level PM2.5 estimation. Based on the re-sidual neural network (ResNet), this study used random oversampling to construct a data-augmented deep re-sidual learning model (AugResNet). In addition, a deep residual neural network model with a particle size constraint was established and called ConResNet. Then, the data augmentation and particle size constraint were incorporated into the deep residual neural network model (denoted as Aug_ConResNet). We evaluated the above four models across China with the 10-fold site-based cross-validation approach. In terms of the estimation ac-curacy for high PM2.5 concentrations (> 75 mu g/m3), AugResNet (R2 = 0.813, RMSE = 19.152 mu g/m3), ConResNet (R2 = 0.796, RMSE = 20.841 mu g/m(3)) and Aug_ConResNet (R2 = 0.820, RMSE = 18.810 mu g/m3) outperformed ResNet (R2 = 0.780, RMSE = 21.628 mu g/m3). Results showed that the data augmentation and particle size constraint alleviated the problem of high-value underestimation and improved the estimation accuracy for high PM2.5 concentrations. The accurate estimation of high PM2.5 concentrations has important application potential for remote sensing monitoring of polluted weather.

学科与关键词

关键词
PM2.5 remote sensingHigh-value underestimationDeep learningData augmentationParticle size constraint
相关
加载中…
  • New interpretable deep learning model to monitor real-time PM2.5 concentrations from satellite data
    Yan, Xing; Zang, Zhou; Luo, Nana; Jiang, Yize; Li, Zhanqing*ENVIRONMENT INTERNATIONAL, 2020, 144: 106060.被引 73
  • Global synthesis of two decades of research on improving PM2.5 estimation models from remote sensing and data science perspectives
    Bai, Kaixu*; Li, Ke; Sun, Yibing; Wu, Lv; Zhang, Ying; Chang, Ni-Bin; Li, Zhengqiang*Earth-Science Reviews, 2023, 241: 104461.被引 10
  • Densely connected convolutional networks with attention long short-term memory for estimating PM2.5 values from images
    Zhang, Bo; Geng, Ziyao; Zhang, Hanwen; Pan, Jianguo*JOURNAL OF CLEANER PRODUCTION, 2022, 333: 130101.被引 8
查看全部 →

发表信息

论文状态
公开发表
期刊
Atmospheric Environment
发表日期
2022-10-15
卷期
第 287 卷
页码
119282
DOI
10.1016/j.atmosenv.2022.119282
语言
ENGLISH
期刊等级
3区
成果收录
SCIE
资助标注
Guangdong Basic and Applied Basic Research Foundation [2022A1515010492]; State Key Laboratory of Resources and Environmental Information System

作者

  • Y

    Yin, Shunchao (第一作者)

    sun yat sen univ

  • L

    Li, Tongwen (通讯)

    sun yat sen univ

查看全部 4 位作者收起作者
  • C

    Cheng, Xiao (通讯)

    sun yat sen univ

  • W

    Wu, Jingan

    sun yat sen univ

登录解锁更多功能

  • 收藏这篇,随时在“我的 · 收藏”找回
  • 认领你的成果,生成推广素材
  • 关注作者获取最新成果
  • 加入项目组、协同写作
注册 / 登录

产品服务

  • 科研之友
  • 创新城
  • 科创云

服务支持

  • 帮助中心
  • 隐私政策
  • 服务条款

联系方式

在线客服:【】
客服热线:400-1616-289
电子邮箱:support@scholarmate.com

关注或下载科研之友

微信二维码
微信公众号
客户端下载二维码
下载客户端
科研成果科研人员科研机构科研动态爱瑞思软件

©2026 深圳市科研之友网络服务有限公司

公安备案图标粤公网安备 44030502000213
粤ICP备 16046710 号粤B2-20110417