Preprints
https://doi.org/10.20944/preprints202312.1052.v1
https://doi.org/10.20944/preprints202312.1052.v1
08 Mar 2024
 | 08 Mar 2024
Status: this preprint is open for discussion and under review for Atmospheric Measurement Techniques (AMT).

Innovative Cloud Quantification: Deep Learning Classification and Finite Element Clustering for Ground-Based All Sky Imaging

Jingxuan Luo, Yubing Pan, Debin Su, Jinhua Zhong, Lingxiao Wu, Wei Zhao, Xiaoru Hu, Zhengchao Qi, Daren Lu, and Yinan Wang

Abstract. Accurate cloud quantification is essential in climate change research. In this work, we construct an automated computer vision framework by synergistically incorporating deep neural networks and finite element clustering to achieve robust whole sky image-based cloud classification, adaptive segmentation, and recognition under intricate illumination dynamics. A bespoke YOLOv8 architecture attains over 95 % categorical precision across four archetypal cloud varieties curated from extensive annual observations (2020) at a Tibetan highland station. Tailor-made segmentation strategies adapted to distinct cloud configurations, allied with illumination-invariant image enhancement algorithms, effectively eliminate solar interference and substantially boost quantitative performance even in illumination-adverse analysis scenarios. In comparison to traditional NRBR threshold analysis methods, the cloud quantification accuracy computed within the framework of this paper exhibits an improvement of nearly 20 %. Collectively, the methodological innovations provide an advanced solution to markedly escalate cloud quantification precision levels imperative for climate change research, while offering a paradigm for cloud analytics transferable to various meteorological stations.

Jingxuan Luo, Yubing Pan, Debin Su, Jinhua Zhong, Lingxiao Wu, Wei Zhao, Xiaoru Hu, Zhengchao Qi, Daren Lu, and Yinan Wang

Status: open (until 14 Apr 2024)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2024-678', Anonymous Referee #1, 25 Mar 2024 reply
    • AC1: 'Reply on RC1', Yinan Wang, 03 Apr 2024 reply
  • RC2: 'Comment on egusphere-2024-678', Anonymous Referee #2, 27 Mar 2024 reply
    • AC2: 'Reply on RC2', Yinan Wang, 03 Apr 2024 reply
  • RC3: 'Comment on egusphere-2024-678', Anonymous Referee #3, 08 Apr 2024 reply
  • RC4: 'Comment on egusphere-2024-678', Anonymous Referee #4, 12 Apr 2024 reply
Jingxuan Luo, Yubing Pan, Debin Su, Jinhua Zhong, Lingxiao Wu, Wei Zhao, Xiaoru Hu, Zhengchao Qi, Daren Lu, and Yinan Wang
Jingxuan Luo, Yubing Pan, Debin Su, Jinhua Zhong, Lingxiao Wu, Wei Zhao, Xiaoru Hu, Zhengchao Qi, Daren Lu, and Yinan Wang

Viewed

Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.

Total article views: 117 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
117 0 0 117 0 0
  • HTML: 117
  • PDF: 0
  • XML: 0
  • Total: 117
  • BibTeX: 0
  • EndNote: 0
Views and downloads (calculated since 08 Mar 2024)
Cumulative views and downloads (calculated since 08 Mar 2024)

Viewed (geographical distribution)

Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.

Total article views: 128 (including HTML, PDF, and XML) Thereof 128 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 13 Apr 2024
Download
Short summary
Accurate cloud quantification is essential for climate research. We developed a novel computer vision framework using deep neural networks and clustering algorithms for cloud classification and segmentation from ground-based all-sky images. Trained on year-round observations, our model achieved over 95 % accuracy for four cloud types. This framework enhances quantitative analysis, supporting climate studies by providing reliable cloud data.