full
search.png

Search

close.png

Cancel

arrow
Volume 31, Issue 1
A Deep Spatio-Temporal Forecasting Model for Multi-Site Weather Prediction Post-Processing

Wenjia Kong, Haochen Li, Chen Yu, Jiangjiang Xia, Yanyan Kang & Pingwen Zhang

DOI: 10.4208/cicp.OA-2020-0158

Commun. Comput. Phys., 31 (2022), pp. 131-153.

Published online: 2021-12

Preview Full PDF 3535 43539

Cited by

google scholar semantic scholar
Export citation
  • Abstract

In this paper, we propose a deep spatio-temporal forecasting model (DeepSTF) for multi-site weather prediction post-processing by using both temporal and spatial information. In our proposed framework, the spatio-temporal information is modeled by a CNN (convolutional neural network) module and an encoder-decoder structure with the attention mechanism. The novelty of our work lies in that our model takes full account of temporal and spatial characteristics and obtain forecasts of multiple meteorological stations simultaneously by using the same framework. We apply the DeepSTF model to short-term weather prediction at 226 meteorological stations in Beijing. It significantly improves the short-term forecasts compared to other widely-used benchmark models including the Model Output Statistics method. In order to evaluate the uncertainty of the model parameters, we estimate the confidence intervals by bootstrapping. The results show that the prediction accuracy of the DeepSTF model has strong stability. Finally, we evaluate the impact of seasonal changes and topographical differences on the accuracy of the model predictions. The results indicate that our proposed model has high prediction accuracy.

  • Keywords

Weather forecasting, post-processing, spatio-temporal modeling, deep learning.

  • AMS Subject Headings

62P12, 86A10, 93B15, 97M10

  • Copyright

COPYRIGHT: © Global Science Press

  • Email address

yuchen1995@pku.edu.cn (Chen Yu)

  • BibTex
  • RIS
  • TXT
@Article{CiCP-31-131, author = {Wenjia Kong , Haochen Li , Yu , ChenJiangjiang Xia , Yanyan Kang , and Zhang , Pingwen}, title = {A Deep Spatio-Temporal Forecasting Model for Multi-Site Weather Prediction Post-Processing}, journal = {Communications in Computational Physics}, year = {2021}, volume = {31}, number = {1}, pages = {131--153}, abstract = {

In this paper, we propose a deep spatio-temporal forecasting model (DeepSTF) for multi-site weather prediction post-processing by using both temporal and spatial information. In our proposed framework, the spatio-temporal information is modeled by a CNN (convolutional neural network) module and an encoder-decoder structure with the attention mechanism. The novelty of our work lies in that our model takes full account of temporal and spatial characteristics and obtain forecasts of multiple meteorological stations simultaneously by using the same framework. We apply the DeepSTF model to short-term weather prediction at 226 meteorological stations in Beijing. It significantly improves the short-term forecasts compared to other widely-used benchmark models including the Model Output Statistics method. In order to evaluate the uncertainty of the model parameters, we estimate the confidence intervals by bootstrapping. The results show that the prediction accuracy of the DeepSTF model has strong stability. Finally, we evaluate the impact of seasonal changes and topographical differences on the accuracy of the model predictions. The results indicate that our proposed model has high prediction accuracy.

}, issn = {1991-7120}, doi = {https://doi.org/10.4208/cicp.OA-2020-0158}, url = {http://global-sci.org/intro/article_detail/cicp/20020.html} }
TY - JOUR T1 - A Deep Spatio-Temporal Forecasting Model for Multi-Site Weather Prediction Post-Processing AU - Wenjia Kong , AU - Haochen Li , AU - Yu , Chen AU - Jiangjiang Xia , AU - Yanyan Kang , AU - Zhang , Pingwen JO - Communications in Computational Physics VL - 1 SP - 131 EP - 153 PY - 2021 DA - 2021/12 SN - 31 DO - http://doi.org/10.4208/cicp.OA-2020-0158 UR - https://global-sci.org/intro/article_detail/cicp/20020.html KW - Weather forecasting, post-processing, spatio-temporal modeling, deep learning. AB -

In this paper, we propose a deep spatio-temporal forecasting model (DeepSTF) for multi-site weather prediction post-processing by using both temporal and spatial information. In our proposed framework, the spatio-temporal information is modeled by a CNN (convolutional neural network) module and an encoder-decoder structure with the attention mechanism. The novelty of our work lies in that our model takes full account of temporal and spatial characteristics and obtain forecasts of multiple meteorological stations simultaneously by using the same framework. We apply the DeepSTF model to short-term weather prediction at 226 meteorological stations in Beijing. It significantly improves the short-term forecasts compared to other widely-used benchmark models including the Model Output Statistics method. In order to evaluate the uncertainty of the model parameters, we estimate the confidence intervals by bootstrapping. The results show that the prediction accuracy of the DeepSTF model has strong stability. Finally, we evaluate the impact of seasonal changes and topographical differences on the accuracy of the model predictions. The results indicate that our proposed model has high prediction accuracy.

Wenjia Kong , Haochen Li , Yu , ChenJiangjiang Xia , Yanyan Kang , and Zhang , Pingwen. (2021). A Deep Spatio-Temporal Forecasting Model for Multi-Site Weather Prediction Post-Processing. Communications in Computational Physics. 31 (1). 131-153. doi:10.4208/cicp.OA-2020-0158
Copy to clipboard
BibteX RIS TXT
The citation has been copied to your clipboard