Preprints
https://doi.org/10.5194/essd-2024-137
https://doi.org/10.5194/essd-2024-137
24 Jun 2024
 | 24 Jun 2024
Status: this preprint is currently under review for the journal ESSD.

MDG625: A daily high-resolution meteorological dataset derived by geopotential-guided attention network in Asia (1940–2023)

Zijiang Song, Zhixiang Cheng, Yuying Li, Shanshan Yu, Xiaowen Zhang, Lina Yuan, and Min Liu

Abstract. The long-term and reliable meteorological reanalysis dataset with high spatial-temporal resolution is crucial for various hydrological and meteorological applications, especially in regions or periods with scarce in situ observations and with limited open-access data. Based on the ERA5 (produced by the European Centre for Medium-Range Weather Forecasts, 0.25°×0.25°, since 1940) and CLDAS (China Meteorological Administration Land Data Assimilation System, 0.0625°×0.0625°, since 2008), we proposed a novel downscaling method Geopotential-guide Attention Network (GeoAN) leveraging the high spatial resolution of CLDAS and the extended historical coverage of ERA5 and produced the daily multi-variable (2 m temperature, surface pressure, and 10 m wind speed) meteorological dataset MDG625 (Song et al., 2024). MDG625 (0.0625° Meteorological Dataset derived by GeoAN) covers most of Asia from 0.125° S to 64.875° N and 60.125° E to 160.125° E since 1940. Compared with other downscaling methods, GeoAN shows better performance with the R2 (2 m temperature, surface pressure, and 10 m wind speed reached 0.990, 0.998, and 0.781, respectively). MDG625 demonstrates superior continuity and consistency from both spatial and temporal perspectives. We anticipate that this GeoAN method and this dataset MDG625 will aid in climate studies of Asia and will contribute to improving the accuracy of reanalysis products from the 1940s. The dataset (Song et al., 2024) is presented in https://doi.org/10.57760/sciencedb.17408 and the code can be found in https://github.com/songzijiang/GeoAN.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.
Zijiang Song, Zhixiang Cheng, Yuying Li, Shanshan Yu, Xiaowen Zhang, Lina Yuan, and Min Liu

Status: open (until 08 Dec 2024)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on essd-2024-137', Anonymous Referee #1, 12 Jul 2024 reply
    • AC1: 'Reply on RC1', Zijiang Song, 19 Jul 2024 reply
Zijiang Song, Zhixiang Cheng, Yuying Li, Shanshan Yu, Xiaowen Zhang, Lina Yuan, and Min Liu

Data sets

MDG625: Meteorological Dataset with 0.0625° resolution produced by GeoAN Zijiang Song et al. https://doi.org/10.57760/sciencedb.17408

Model code and software

Code of GeoAN Zijiang Song https://github.com/songzijiang/GeoAN

Zijiang Song, Zhixiang Cheng, Yuying Li, Shanshan Yu, Xiaowen Zhang, Lina Yuan, and Min Liu

Viewed

Total article views: 525 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
435 63 27 525 16 17
  • HTML: 435
  • PDF: 63
  • XML: 27
  • Total: 525
  • BibTeX: 16
  • EndNote: 17
Views and downloads (calculated since 24 Jun 2024)
Cumulative views and downloads (calculated since 24 Jun 2024)

Viewed (geographical distribution)

Total article views: 508 (including HTML, PDF, and XML) Thereof 508 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 06 Nov 2024
Download
Short summary
It is hard to access long-time series and high-resolution meteorological data for historical years. In this paper, we proposed a geopotential-guide attention network (GeoAN) for downscaling, which can produce high-resolution data from the given low-resolution data. Quantitative and visual comparisons reveal our GeoAN producing better results in most metrics. Using GeoAN, historical meteorological data called MDG625 has been produced daily since 1940.
Altmetrics