Jump to main content
US EPA
United States Environmental Protection Agency
Search
Search
Main menu
Environmental Topics
Laws & Regulations
About EPA
Health & Environmental Research Online (HERO)
Contact Us
Print
Feedback
Export to File
Search:
This record has one attached file:
Add More Files
Attach File(s):
Display Name for File*:
Save
Citation
Tags
HERO ID
7723515
Reference Type
Journal Article
Title
Forecasting PM2.5 using hybrid graph convolution-based model considering dynamic wind-field to offer the benefit of spatial interpretability
Author(s)
Zhou, H; Zhang, F; Du, Z; Liu, R
Year
2021
Is Peer Reviewed?
Yes
Journal
Environmental Pollution
ISSN:
0269-7491
EISSN:
1873-6424
Volume
273
Page Numbers
116473
Language
English
PMID
33503566
DOI
10.1016/j.envpol.2021.116473
Web of Science Id
WOS:000625376600058
Abstract
Air pollution is a complex process and is affected by meteorological conditions and other chemical components. Numerous studies have demonstrated that data-driven spatio-temporal prediction models of PM2.5 concentration are comparable with the model-driven model. However, data-driven models are usually depending on the statistical correlation between PM2.5 and other factors and have challenges in dealing with causality in complex systems. In this paper, we argue that domain knowledge should be incorporated into data-driven models to enhance prediction accuracy and make the model more physically realistic. We focus on the influence of dynamic wind-field on PM2.5 concentration distribution and fuse the pollution diffusion distance with the deep learning model based on a wind-field surface. In order to model spatial dependence between monitoring stations, which is dynamic and anisotropic because of the wind-field, we proposed a hybrid deep learning framework, dynamic directed spatio-temporal graph convolution networks (DD-STGCN). It expanded the ability to deal with space-time prediction in the continuous and dynamic wind-field. We used a directed graph time-series to describe the vertex state and topological relationship between vertices and replaced traditional Euclidean distance with wind-field diffusion distance to describe the proximity relationship between vertices. Our experiment results demonstrated that the DD-STGCN model achieved a better prediction ability than LSTM, GC-LSTM, and STGCN models. Compared to the best comparison model, MAPE, MAE, and RMSE were improved by 10.2%, 9.7%, and 9.6% in 12 h on an average, respectively. The performance of our model was further tested during a haze period. In the case that two models both considered the effect of wind, compared with the pure data-driven model, our model performed better in prediction distribution and showed the benefit of spatial interpretability provided by domain knowledge.
Keywords
PM2.5 concentration forecast; Domain knowledge; Dynamic wind-field; Graph convolution network; Temporal convolution network
Tags
•
LitSearch-NOx (2024)
Forward Citation Search
Epidemiology
Results
Reproductive
PubMed
WoS
Exposure
Results
Error Impacts
PubMed
WoS
Error Sources
WoS
Home
Learn about HERO
Using HERO
Search HERO
Projects in HERO
Risk Assessment
Transparency & Integrity