Preprints
https://doi.org/10.5194/gmd-2022-272
https://doi.org/10.5194/gmd-2022-272
Submitted as: development and technical paper
09 Jan 2023
Submitted as: development and technical paper | 09 Jan 2023
Status: this preprint is currently under review for the journal GMD.

Convective Gusts Nowcasting Based on Radar Reflectivity and a Deep Learning Algorithm

Haixia Xiao1, Yaqiang Wang2,, Yu Zheng1,, Yuanyuan Zheng1, Xiaoran Zhuang1,3, Hongyan Wang2, and Mei Gao2 Haixia Xiao et al.
  • 1Key Laboratory of Transportation Meteorology of China Meteorological Administration, Nanjing Joint Institute for Atmospheric Sciences, Nanjing 210041, China
  • 2Chinese Academy of Meteorological Sciences, Beijing 100081, China
  • 3Jiangsu Meteorological Observatory, Nanjing 210041, China
  • These authors contributed equally to this work.

Abstract. Convective wind gusts (CGs) are usually related to thunderstorms, and they may cause great structural damage and serious hazards, such as train derailment, service interruption, and building collapse. Due to the small-scale and nonstationary nature of CGs, reliable CGs nowcasting with high spatial and temporal resolutions has remained unattainable. In this study, a novel nowcasting model based on deep learning – namely, CGsNet – is developed for 0–2 h of quantitative CGs nowcasting, first achieving minute-kilometer-level forecasts. CGsNet is a physics-constrained model established by training on large corpora of average surface wind speed (ASWS) and radar observations, it can produce realistic and spatiotemporally consistent ASWS predictions in CGs events. By combining the gust factor (1.77, the ratio of the observed peak wind gust speed (PWGS) to the ASWS) with the ASWS predictions, the PWGS forecasts are estimated with a spatial resolution of 0.01° × 0.01° and a 6-minute temporal resolution. CGsNet is shown to be effective, and it has an essential advantage in learning the spatiotemporal features of CGs. In addition, quantitative evaluation experiments indicate that CGsNet exhibits higher generalization performance for CGs than the traditional nowcasting method based on numerical weather prediction models. CGs nowcasting technology can be applied to provide real-time quantitative CGs forecasts and alerts the damaging wind events in meteorological services.

Haixia Xiao et al.

Status: open (until 06 Mar 2023)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse

Haixia Xiao et al.

Data sets

Convective Gusts Datasets( radar reflectivity and wind observations) Haixia Xiao, Yaqiang Wang, Yu Zheng, Yuanyuan Zheng, Xiaoran Zhuang, Hongyan Wang, and Mei Gao https://doi.org/10.7910/DVN/PIZU7V

Model code and software

The code of CGsNet model Haixia Xiao, Yaqiang Wang, Yu Zheng, Yuanyuan Zheng, Xiaoran Zhuang, Hongyan Wang, and Mei Gao https://doi.org/10.7910/DVN/PIZU7V

Haixia Xiao et al.

Viewed

Total article views: 194 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
162 29 3 194 0 1
  • HTML: 162
  • PDF: 29
  • XML: 3
  • Total: 194
  • BibTeX: 0
  • EndNote: 1
Views and downloads (calculated since 09 Jan 2023)
Cumulative views and downloads (calculated since 09 Jan 2023)

Viewed (geographical distribution)

Total article views: 183 (including HTML, PDF, and XML) Thereof 183 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 28 Jan 2023
Download
Short summary
Due to the small-scale and nonstationary nature of convective wind gusts (CGs), reliable CGs nowcasting has remained unattainable. Here, we developed a deep learning model – namely CGsNet – for 0–2 hours of quantitative CGs nowcasting, first achieving minute-kilometer-level forecasts. Based on CGsNet model, the average surface wind speed (ASWS) and peak wind gust speed (PWGS) predictions are obtained. Experiments indicate that CGsNet exhibits higher accuracy than the traditional method.