Skip to main content
Log in

Quantification of the concrete freeze–thaw environment across the Qinghai–Tibet Plateau based on machine learning algorithms

  • Original Article
  • Published:
Journal of Mountain Science Aims and scope Submit manuscript

Abstract

The reasonable quantification of the concrete freezing environment on the Qinghai–Tibet Plateau (QTP) is the primary issue in frost resistant concrete design, which is one of the challenges that the QTP engineering managers should take into account. In this paper, we propose a more realistic method to calculate the number of concrete freeze-thaw cycles (NFTCs) on the QTP. The calculated results show that the NFTCs increase as the altitude of the meteorological station increases with the average NFTCs being 208.7. Four machine learning methods, i.e., the random forest (RF) model, generalized boosting method (GBM), generalized linear model (GLM), and generalized additive model (GAM), are used to fit the NFTCs. The root mean square error (RMSE) values of the RF, GBM, GLM, and GAM are 32.3, 4.3, 247.9, and 161.3, respectively. The R2 values of the RF, GBM, GLM, and GAM are 0.93, 0.99, 0.48, and 0.66, respectively. The GBM method performs the best compared to the other three methods, which was shown by the results of RMSE and R2 values. The quantitative results from the GBM method indicate that the lowest, medium, and highest NFTC values are distributed in the northern, central, and southern parts of the QTP, respectively. The annual NFTCs in the QTP region are mainly concentrated at 160 and above, and the average NFTCs is 200 across the QTP. Our results can provide scientific guidance and a theoretical basis for the freezing resistance design of concrete in various projects on the QTP.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

Availability of Data/Materials: The MERRA2 dataset was obtained from https://gmao.gsfc.nasa.gov/reanalysis/MERRA-2/data_access/. The China Meteorological Forcing Dataset (CMFD) (http://www.tpedatabase.cn/)The datasets generated during this study are available from the corresponding author upon reasonable request.

References

Download references

Acknowledgments

The authors would like to thank all the public bodies that have made the data available, through their online digital archives. This work was jointly supported by Shandong Provincial Natural Science Foundation (grant number: ZR2023MD036), Key Research and Development Project in Shandong Province (grant number: 2019GGX101064), project for excellent youth foundation of the innovation teacher team, Shandong (grant number: 2022KJ310). The authors also thank the National Tibetan Plateau Data Center (u]http://data.tpdc.ac.cn) and the Global Modelling and Assimilation Office for providing the data for this study.

Author information

Authors and Affiliations

Authors

Contributions

QIN Yanhui: Material preparation, Data collection, Writing-original draft. MA Haoyuan: Data analysis, Visualization. ZHANG Lele: Writing-review & editing. YIN Jinshuai: Literature search, Conceptualization, Li Shuo: Supervision, Writing-review.

Corresponding author

Correspondence to Yanhui Qin.

Ethics declarations

Conflict of Interest: The authors declare no conflict of interest.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Qin, Y., Ma, H., Zhang, L. et al. Quantification of the concrete freeze–thaw environment across the Qinghai–Tibet Plateau based on machine learning algorithms. J. Mt. Sci. 21, 322–334 (2024). https://doi.org/10.1007/s11629-023-8389-7

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11629-023-8389-7

Keywords

Navigation