Real-time mapping of gapless 24-hour surface PM_(10) in China  

作　　者：Xutao Zhang Ke Gui Hengheng Zhao Nanxuan Shang Zhaoliang Zeng Wenrui Yao Lei Li Yu Zheng Hujia Zhao Yurun Liu Yucong Miao Yue Peng Ye Fei Fugang Li Baoxin Li Hong Wang Zhili Wang Yaqiang Wang Huizheng Che Xiaoye Zhang 

机构地区：[1]State Key Laboratory of Severe Weather and Key Laboratory of Atmospheric Chemistry of China Meteorological Administration,Chinese Academy of Meteorological Sciences,Beijing 100081,China [2]Institute of Artificial Intelligence for Meteorology,Chinese Academy of Meteorological Sciences,Beijing 100081,China [3]Institute of Atmospheric Environment,China Meteorological Administration,Shenyang 110166,China [4]Plateau Atmospheric and Environment Key Laboratory of Sichuan Province,College of Atmosphere Sciences,Chengdu University of Information Technology,Chengdu 610225,China [5]National Meteorological Information Center,Beijing 100081,China [6]China Global Atmosphere Watch Baseline Observatory,Xining 810001,China [7]Greenhouse Gas and Carbon Neutral Key Laboratory of Qinghai Province,Xining 810001,China

出　　处：《National Science Review》2025年第2期111-123,共13页国家科学评论(英文版)

基　　金：supported by the National Key Research and Development Program of China(2023YFC3706305);the Science and Technology Plan Project of the China Meteorological Administration(CMAJBGS202325);the National Natural Science Foundation of China project(42175153 and 42030608);the Youth Innovation Team of China Meteorological Administration(CMA2024QN13);the Third Xinjiang Scientific Expedition Program(2022xjkk0903);the Basic Research Fund of CAMS(2023Z021).

摘　　要：Large-scale mapping of surface coarse particulate matter(PM_(10))concentration remains a key focus for air quality monitoring.Satellite aerosol optical depth(AOD)-based data fusion approaches decouple the non-linear AOD-PM_(10) relationship,enabling high-resolution PM_(10) data acquisition,but are limited by spatial incompleteness and the absence of nighttime data.Here,a gridded visibility-based real-time surface PM_(10) retrieval(RT-SPMR)framework for China is introduced,addressing the gap in seamless hourly PM_(10) data within the 24-hour cycle.This framework utilizes multisource data inputs and dynamically updated machine-learning models to produce 6.25-km gridded 24-hour PM_(10) data.Cross-validation showed that the RT-SPMR model’s daily retrieval accuracy surpassed prior studies.Additionally,through rolling iterative validation experiments,the model exhibited strong generalization capability and stability,demonstrating its suitability for operational deployment.Taking a record-breaking dust storm as an example,the model proved effective in tracking the fine-scale evolution of the dust intrusion process,especially in under-observed areas.Consequently,the operational RT-SPMR framework provides comprehensive real-time capabi lity for monitoring PM_(10) pol lution in China,and has the potential to improve the accuracy of dust storm forecasting models by enhancing the PM_(10) initial field.

关 键 词：PM_(10) real time seamless retrieval interpretable machine learning dust storm 

分 类 号：X51[环境科学与工程—环境工程]