Home Journals Articles Updates About
CN
image
收藏切换
Research on pollen concentration prediction technology in Beijing based on the fusion of machine learning and time series algorithms
收藏切换
Tian-feng DENG1, Zi-ming LI2, *, Shao-bo ZHANG1, **, Zhao-bin SUN3
China Environmental Science | 2025, 45(11) : 6007 - 6019
Less
收藏切换
China Environmental Science | 2025, 45(11): 6007-6019
• Air Pollution Control •
Research on pollen concentration prediction technology in Beijing based on the fusion of machine learning and time series algorithms
Full
Tian-feng DENG1, Zi-ming LI2, *, Shao-bo ZHANG1, **, Zhao-bin SUN3
Affiliations
  • 1.Chengdu Plain Urban Meteorology and Environment Observation and Research Station of Sichuan Province, Sichuan Provincial Engineering Research Center for Meteorological Disaster Prediction and Early Warning, Plateau Atmosphere and Environment Key Laboratory of Sichuan Province, School of Atmospheric Sciences, Chengdu University of Information Technology, Chengdu 610225, China
  • 2.Beijing Meteorological Station, Beijing 100089, China
  • 3.Center for Meteorological Impact and Risk Research, Chinese Academy of Meteorological Sciences, Beijing 100081, China
Published: 2025-11-20
Outline
收藏切换

In this paper, a time-series simulation prediction study of pollen concentration in Shunyi District, Beijing, China, for the period of 2018~2024 was carried out based on machine learning and time-series algorithms. The study integrates traditional statistical models, machine learning models and deep learning models, and introduces 36 meteorological elements and their variables as covariates to improve the prediction accuracy. The results show that AutoGluon-TimeSeries (AG-TS) significantly reduces the risk of overfitting through model integration, and performs well in short-term prediction (1day), with DLinear model having the best overall performance with R=0.68 and RMSE=206.5. In contrast, some models overestimated or underestimated the predictions. It was found that the pollen concentration showed double peaks in spring (March~May) and fall (August~October), and the correlation with meteorological variables reflected the complex effects of meteorological conditions and vegetation phenology. In addition, the modeling performance declined with the extension of prediction time, e.g., the R-value of DLinear decreased by 18% and the RMSE increased by 15% in the 3-day prediction. By eliminating the inefficient models and optimizing the remaining models, the direction of model optimization and scientific support for pollen allergy early warning were provided.

time series  /  forecasting  /  pollen concentration  /  machine learning  /  model ensemble
邓田丰, 李梓铭, 张少波, 孙兆彬. 基于机器学习和时间序列算法融合的北京地区花粉浓度预测技术研究. 中国环境科学, 2025 , 45 (11) : 6007 -6019 .
Tian-feng DENG, Zi-ming LI, Shao-bo ZHANG, Zhao-bin SUN. Research on pollen concentration prediction technology in Beijing based on the fusion of machine learning and time series algorithms[J]. China Environmental Science, 2025 , 45 (11) : 6007 -6019 .
Appendix
Less
Year 2025 volume 45 Issue 11
70
19
Cite this Article
BibTeX
Article Info
  • Receive Date：2025-04-09
  • Online Date：2026-10-08
  • Published：2025-11-20
Article Data
Affiliations
History
  • Received：2025-04-09
Funding
Affiliations
    1.Chengdu Plain Urban Meteorology and Environment Observation and Research Station of Sichuan Province, Sichuan Provincial Engineering Research Center for Meteorological Disaster Prediction and Early Warning, Plateau Atmosphere and Environment Key Laboratory of Sichuan Province, School of Atmospheric Sciences, Chengdu University of Information Technology, Chengdu 610225, China
    2.Beijing Meteorological Station, Beijing 100089, China
    3.Center for Meteorological Impact and Risk Research, Chinese Academy of Meteorological Sciences, Beijing 100081, China
References
Share
https://castjournals.cast.org.cn/joweb/zghjkx/EN/
Share to
QR

Scan QR to access full text

Cite this article
BibTeX
Citations
表12种不同金属材料的力学参数
科
Family		 属数
Number of
genus		 种数
Number of
species		 占总种数比例
Percentage of
total species (%)		 属
Genus		 种数
Number of
species		 占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏
  • BibTeX
  • EndNote
  • RefWorks
  • TxT
Links
Articles
Latest Articles
Most Read
Collections
Updates
Events
News
Multimedia
About
About Us
Contact
No. 86 Xueyuan South Road, Haidian District, Beijing
100081
010-62662699(Operation and Technical Support)
qkjq@cast.org.cn

Copyright © 2025 China Association for Science and Technology. All rights reserved. For all open access content, the relevant licensing terms apply.
Sponsored by the Office of the Leading Group for Cybersecurity and Informatization of CAST, and supported by Science and Technology Review Publishing House