Geoscientific Model Development
Geoscientific Model Development
GMD
Articles | Volume 15, issue 22
Articles | Volume 15, issue 22
https://doi.org/10.5194/gmd-15-8439-2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/gmd-15-8439-2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.
Articles | Volume 15, issue 22
Model description paper
 | 
21 Nov 2022
Model description paper |  | 21 Nov 2022

Development of an LSTM broadcasting deep-learning framework for regional air pollution forecast improvement

Haochen Sun, Jimmy C. H. Fung, Yiang Chen, Zhenning Li, Dehao Yuan, Wanying Chen, and Xingcheng Lu

Supplement

https://doi.org/10.5194/gmd-15-8439-2022-supplement

Data sets

Ground obeservation data (meteorological factors and air pollution) in Greater Bay Area, 2015-2021 Haochen Sun, Jimmy Chi Hung Fung, Yiang Chen, Zhenning Li, Dehao Yuan, Wanying Chen, and Xingcheng Lu https://doi.org/10.5281/zenodo.6598377

Prediction of the broadcasting model and various baselines Haochen Sun, Jimmy Chi Hung Fung, Yiang Chen, Zhenning Li, Dehao Yuan, Wanying Chen, and Xingcheng Lu https://doi.org/10.5281/zenodo.6833673

Deep learning models in the study "Development of an LSTM-Broadcasting deep-learning framework for regional air pollution forecast improvement" Haochen Sun, Jimmy Chi Hung Fung, Yiang Chen, Zhenning Li, Dehao Yuan, Wanying Chen, and Xingcheng Lu https://doi.org/10.5281/zenodo.6827585

Processed ground observation and WRF-CAMQ data for Greater Bay Area, 2015-2021 Haochen Sun, Jimmy Chi Hung Fung, Yiang Chen, Zhenning Li, Dehao Yuan, Wanying Chen, and Xingcheng Lu https://doi.org/10.5281/zenodo.6601173

Model code and software

jvhs0706/regional-forecast-new: GMD paper code Jeff Haochen Sun https://doi.org/10.5281/zenodo.7019243

CMAQ (Version 5.0.2) United States Environmental Protection Agency https://doi.org/10.5281/zenodo.1079898

Download
Short summary
This study developed a novel deep-learning layer, the broadcasting layer, to build an end-to-end LSTM-based deep-learning model for regional air pollution forecast. By combining the ground observation, WRF-CMAQ simulation, and the broadcasting LSTM deep-learning model, forecast accuracy has been significantly improved when compared to other methods. The broadcasting layer and its variants can also be applied in other research areas to supersede the traditional numerical interpolation methods.
Read more